Analysis of the Interlanguage of Second Language Learners: Implications for the Classroom

Abstract

This chapter provides analysis of the interlanguage of second language learners and the pedagogic implications. It begins with reviews of the concept of interlanguage from the background of cross-linguistic influence, providing analysis of interlanguage of second language (SL) learners, proof of systematic variability of interlanguage use at the syntactic level as well as implications for SL teachers. It discusses the revised interlanguage hypothesis, revealing the shift from the previous concept where interlanguage solely applied to adults acquiring second languages to children who produce interlanguages with socio-linguistic justifications. This study highlights the role of social contexts in interlanguage development. Finally, the study reviews the concept of fossilization, causes of fossilization, and the arguments on inevitability of fossilization in second language learning. Pedagogically, the chapter suggests that interlanguage system can be changed by employing different teaching methods and strategies, creating linguistic and social contexts which are facilitative, and delisting those that inhibit learning and classroom achievements.

1. Introduction

One remarkable area of research into second language acquisition is the nature and properties of the second language learner’s knowledge and use of the target language grammar commonly known as interlanguage. The term “interlanguage”(IL) was first coined by Selinker, L. [1] which refers to the language system created by the foreign language learner based on the linguistic input accessible to him. Selinker proposed that a linguistic system (interlanguage, IL) underlies L2 learner’s language which he produces in his attempt to communicate meaningfully in the target language. This system as he further insinuated, has a structure different from the learner’s native and the target languages and ought to be investigated to associate the psychological processes influencing interlanguage over time. Prior to [1] use of the term, [2] in The significance of learner’s error had a notion similar to interlanguage as “transitional competence” which he explored to describe the linguistic ability of foreign language learners which is deviant from the norm used by native speakers. Later, Nemser [3] and Corder [4] independently used “approximation system” and “idiosyncratic dialect”, respectively to refer to the successive approximation to the target language and the uniqueness of the L2 learner’s language which is particular to that individual. Today, idiosyncratic dialect, approximation system, and interlanguage refer to the same phenomenon. In Ref. [1] used interlanguage to refer to the systematic knowledge of language which is independent of both the learner’s native language and the target language. Recently, [1] (2007, p. 256) defines interlanguage as “the separateness of a second language learner’s system that has a structurally intermediate status between the native language and the target language”. L2 learners’ native and target languages have independent linguistic systems, the learner, at this level of second language development, forms what Brown calls the learner’s self-contained linguistic system which is neither the system of the native language nor the target language. These interim systems are regularly revised by L2 learners to adapt new hypotheses to the target language.

Extensive research into the L2 learner’s interlanguage provides strong evidence that variability is common in the interlanguage grammar of all L2 learners [1, 5]. Selinker [1] argues that interlanguage is characterized by permeability, that is, the learner’ grammar is susceptible to infiltration by his first language and the target language rules or forms. He changes his grammar from time to time by deleting rules, adding rules, and reconstructing the complete rules. The rules according to Al-khresheh [6] are viewed as mental grammar which create IL system. Song [5] best describes interlanguage as a system based on the best attempt of learners to provide order and structure to the linguistic stimuli surrounding them. Through a gradual process of trial and error and hypothesis testing, the learner slowly succeeds in establishing closer approximations to the system used by the native speakers of the target language. Before this stage, the learner must first discover that his output does not match the target language. He will never detect a mismatch between his forms and the target language forms if the target forms are passed through an L1 filter [7, 8]. Interlanguage is constantly changing leading to its second characteristic, dynamic. Elis [9] declares that “L2 learner’s mental grammars are perceived as dynamic and subject to rapid change” p. 352. The dynamism is orchestrated by FL influence, stretching, and over-generation of the TL rules. Song [5] reports that the learner’s interlanguage does not uncoordinatedly and suddenly change from one level to another; rather, it slowly revives the interim system to adapt new hypothesis to the target language system. The process of constant revision and extension of rules, as Song further postulates, is a feature of the inherent instability of interlanguage and its built-in propensity for change. Interlanguage incorporates deviant forms or errors associated with the learner’s current linguistic level as he approaches and approximates a standard of the target language; however, there exists a certain degree of uniformity in application of rules by L2 learners. That brings us to the third characteristic; interlanguage is systematic. Song [5] reports that in spite of the instability of interlanguage, it is possible to identify the rule-based nature of the learner’s interlanguage, that is, the learner follows predictable application of the rules of the target language in his efforts to communicate meaning. Wang & Fan [10] confirm that the L2 learner bases their performance skills on the existing rule system of the target language as much as the native speaker basis his competence skills on his internalized knowledge of the L1 system.

Aside from the systematicity of interlanguage, learners’ interlanguage grammars exhibit high degrees of variability resulting from various ways mental systems involved in the construction of interlanguage interact with different types of linguistic input. L2 learners are exposed to explicit instruction and negative feedback- linguistic input, in addition to naturally occurring sentences of the target language [11] which engenders learned linguistic knowledge in them. Each learner’s interlanguage presents alternative rules for performing the same function. The unstable nature of IL system reflects the overlapping stages of development as one variable rule is revised in favor of another. A typical example of interlanguage variability where native speakers would not vary is revealed by [12] who examined the interlanguage of Igbo−English bilinguals who were given Igbo passages to translate to English. Igbo−English bilinguals were found to use attributive and absolute genitives in sentences whereas native speakers would only use absolute genitives, egs, IL (this my brother) standard English (SE) (my brother/this brother of mine). Chidi-Onwuta & Oluikpe [12] observed that Igbo lacks absolute genitives, consequently, the Igbo–English bilingual appeals to prior linguistic knowledge by using attributive genitives in place of absolute genitives not found in Igbo and sometimes, superfluously uses absolute genitives when need be. Their study reveals instances of cross-linguistic influence in L2 learners’ IL. It confirms [13] records that learners of a new language employ a number of strategies in approaching a new language; top in the list is the native language of L2 learners.

2. Cross-linguistic influence (transfer)

Cross-linguistic influence (CLI) or transfer, as may be interchangeably used in this study, is a theory that recognizes the significant role prior experience plays in any learning act [14]. CLI, according to [14], is “the influence resulting from similarities or differences between the target language (TL) and any other language that has already been previously (and perhaps imperfectly) acquired” p. 27. The second language learner appeals to previous knowledge he has had with language to facilitate learning of new experiences. The first language, according to Brown [13], is the obvious set of prior experiences accessible to the L2 learner. Following Brown [13] claim, transfer is imperative in any language learning. Chidi-Onwuta & Oluikpe [12] traced the theory of transfer as being construed within the Behavioral framework as habits generalized from one language to another. Behavioral psychologists claim that transfer is evident in every second language learning [15] and is subdivided into positive and negative. When the first language is similar to the target language to the extent that learning is facilitated by prior experience, positive transfer is in place. Negative transfer, however, is observed when the influences of the first language are inhibitory to the learning of the target language especially when there are remarkable differences between the two languages.

Two different groups of researchers lent their voices to the controversy about the role of transfer; L2 direct UG Access and L2 non-UG Access exponents. A proponent of L2 direct UG Access, [16, 17] reveals that UG operates for L2 learners more or less as a “clean state” free from L1 affectation and uninterrupted by L1 parameterization. Consequently, the L2 learner is uninfluenced by his L1 experiences and linguistic features and structures. To support the claim that L2 learners can build an L2 grammar from the scratch in exactly the way L1 users acquire their first language, empirical studies were independently carried out by Dulay and Burt [18] and Bailey et al. [19] and lately by Cook [20] which reveal that L2 learners from different language backgrounds exhibited remarkable similarities in their acquisition of different grammatical morphemes in the target. Their performances reveal that knowledge of grammatical morphemes and structure dependency did not come from their prior experiences (L1), but rather, from UG principles and parameters of word order [1]. Earlier listed other factors that influence second language learning and shape interlanguage to include: transfer of training, strategies of second language communication, strategies of second language learning, and over-generation.

Transfer of training is a cognitive process where the L2 learner applies rules learned from textbooks or instructors. Sometimes, the L2 learner applies these rules correctly, but at other times, they are superfluously used. For example, when simple past forms of the verb are used to express events that happened in the distant past. Chidi-Onwuta & Oluikpe [12] provide negative evidence of Igbo−English bilinguals who superfluously used forms of the verb and articles to communicate meaning in the TL. For example:

Interlanguage	Standard English
I lived in Abuja before it was made the capital of Nigeria	I had lived in Abuja much early before it became the capital of Nigeria
Okoro gave me some advice that helped me	Okoro offered useful advice to me

The observable errors above are directly traceable to the manner of presentation of the language features in the school course. In a foreign language teaching setting where the major source of comprehensible input in English is the teaching manual and the teacher, errors traceable to transfer of training would affect learners’ IL. The transfer of training errors above reveals an attempt, according to Oluikpe [21], of the Igbo−English bilingual to grapple with idiosyncratic usage patterns of verb forms and articles (ART) in the English language. These errors showed up in the learners’ IL because Igbo does not have all the verb forms as found in English. Again, there is no ART in Igbo language. In using ART and other verb forms not found in Igbo, the Igbo−English bilingual use those linguistic features superfluously.

Strategies of communication are cognitive processes the L2 learner engages to convey meaning when faced with production problems in the TL. Less proficient L2 learners adopt communication strategies like avoidance, message abandonment, literal translation, and language switching while their more proficient counterparts explore strategies like circumlocution, word coinage, and approximation [22]. Factors that affect the L2 learner’s choice of a particular communication strategy depending on the learner’s personality and context.

When the L2 learners’ IL structures and rules are a result of their identifiable approach to the TL, such approach becomes their strategy for second language learning. Possible learning strategies that shape IL rules according to [23] include hypothesis formation (simplification and inferencing) and hypothesis testing. The developmental nature of interlanguage reveals that L2 learners go through a process of making and testing hypotheses about the target language [1], a process which according to Jolali & Shojaei [24] leads to internalizing the rules. Successful hypotheses become mental constructions that correspond to the rules of the TL while unsuccessful hypotheses are revised and discarded [25]. The implication of this claim is that L2 learners are located on an interlanguage (IL) continuum between their L1 and the target language. Highly successful L2 learners attain commensurate level of competence while less successful L2 learners become fossilized somewhere along the IL continuum [26]. The concept of fossilization will be discussed later in this chapter.

Once a hypothesis is formed, the L2 learner tests this hypothesis in a number of ways which include:

Type of test	Method of test
Receptive	Learners compare hypotheses to second language input
Productive	Learners use a hypothesis to generate language and assess feedback
Metalingual	Learners consult a native speaker or text, etc.
Interactional	Learners make an intentional error to elicit a repair from a native speaker.

([27], p. 33).

Another cognitive process that influences IL is over-generalization. In Ref. [28] defines over-generalization as “the use of previously available strategies in new situations” p. 174. This process is widely observed in child language acquisition, and it is termed a developmental process. At this level, the acquirer or the L2 learner shows evidence of mastering a rule but does not take cognizance of exceptions to the rule. For example, the learner knows the past tense marker ‘ed’ is used to express past actions with verbs, without conscious observance of regular and irregular verbs alike: washed, danced, *drinked, *goed, *teached. Chidi-Onwuta & Oluikpe [12] provide negative evidence of adult Igbo−English bilinguals engaged in writing production who instanced over-generalization in their IL:

IL	SE
Staffs at department of Mass communication	The staff of the department of Mass communication
Those advices you gave me	The piece of advice which you offered me

Over-generalization is a deviation from the TL rules which may present in form of addition of linguistic units or omission of required features. Overgeneralization may result in simplification of the target language rules; in that case, the L2 learner is seen reducing the linguistic burden in the TL. The omission of the third person simple present tense marker‘s’ in (she eat rice often) is a typical example of simplification of the TL rules in interlanguage. Overgeneralization error reveals evidence of the learner’s progress in TL development; it shows that the learner has learnt some general language rules and waiting to master the exceptions. This claim, however, is contrary to [29] who earlier sees learners who over-generalize as those who have limited exposure to the TL rules and create deviant structures. Firth further believes that overgeneralization illustrates an attempt by the L2 learner to build up hypothesis about the TL from their limited exposure to it, as such, the following utterances can be found among L2 learners:

*I am not liking it

*she must goes to jim every day.

The review above confirms [13] view that L2 learners employ different strategies in approaching the TL, top in the list is appeal to the native language which results in cross-linguistic influence. Generalizations about cross-linguistic influence have been revisited with empirical studies and such studies are used to support the claim of absence of cross-linguistic influence in second language acquisition [12, 20, 30]. For example, some groups of Igbo−English bilinguals were engaged in a study carried out by [12] where they were to apply pluralization as a categorical feature of nouns in their writing and translation. The following negative evidence were noted in their writing:

Interlanguage (IL)	Standard English (SE)
These three sheeps you killed are your own	The three sheep you slaughtered are yours
I do not get the necessary informations I needed early	I was unable to receive in good time information which I needed
Chike broke some furnitures	Chike broke some furniture

The negative evidence above does not reflect cross-linguistic influence. From the perspective of contrastive analysis (CA), both the target language (English) and the learners’ first language (Igbo) have the noun category. Similarly, both languages have noun subcategories (± count and non-count noun). Again, English and Igbo languages recognize the notion of pluralization with minor differences in plural marking. Pluralization is marked morphologically in English and characterized by exceptions while the same is post-nominally marked in Igbo using numerals. When the noun is singular, the numeral “one” is used pre-nominally to mark singularity. Since noun category and pluralization are features found in both English and Igbo languages, the negative evidence above could be explained to have risen as the Igbo-English bilingual attempts to reinterpret the rules in the two languages by invoking the learning principle of generalization. It is further interpreted that the general principle of generalization as a learning process was at work rather than appealing to L1. Chidi-Onwuta & Oluikpe [12] study reveals all errors in the target language as unexplainable from transfer and could not only be induced by cross-linguistic influence.

Another group of researchers who upheld the role prior experiences play in the learning of a new language is supporters of L2 non-UG Access [14, 31]. These scholars claim that at this level of language learning, when the L2 learner has passed the critical period [32], UG ceases to operate for him. He alternates with a surrogate UG made available to him through his L1. The interlanguage at this level is filled with remarkable L1-influenced errors which are characteristics of L1 parameterization. Following [31], this initial stage of second language learning is characterized by transfer of L1 parameter setting onto L2 with accumulation of L1-induced errors which recede as learning progresses. At a later stage when L2 learner engages in strategies to deal with the differing language inputs, such errors drop off. Qaid & Ramamoorthy [33] supports L2 non-UG Access claim of cross-linguistic influence in L2 learning through a study aimed at identifying and classifying the writing errors of Arabic third-year college of education, English major students. Their findings revealed that interlingual errors caused mainly by negative transfer ranked highest (49.42) among other investigated interlingual errors. The frequent errors in writing production of the students were due to interference of Arabic on English especially in the omission of the required elements in the target language due to lack of those linguistic elements in their first language.

McWhinney & Bates [8], from the competition model perspective, provide other evidence of transfer during the process of phonological, syntactic, and lexical learning. They support the claim of massive transfer of L1 features and structures. According to competition model, L2 learners employ various complex learning strategies; when a simple transfer is blocked, the learners get around it by developing a new set of strategies and postulating more complex remappings from L1 to L2. L2 learners, according to the exponents of the competition model, interpret meaning in the target language through appeal to the first language [13].

Transfer or crosslinguistic influence, following the claims by supporters of L2 Non-UG Aceess and exponents of competition model, shapes IL rules. Although transfer is not the only cognitive process involved in the development of IL, there is ample research evidence that it does affect interlanguage rules [12, 31, 33]. Selinker [34] suggests that the cognitive process underlying transfer is interlingual identification. Prior to Selinker [34] claim, [35] understudied the interlingual identification of Spanish and English vowels by sixty native speakers of Spanish in three experiments judging the vowel /i/, /l/, /Ɛ/ and /æ/ in multiple tokens of English words (beat, bit, bet, bat) spoken by ten native speakers of American English. The students were to label each English vowel by circling one of the five letters used to spell the vowel phonemes of Spanish (viz < e>, <a>, <o>, <u>) or by circling “none” if they thought they had heard a vowel not found in Spanish. Their findings revealed that students who spoke English as an L2 used the “none” label more often than did Spanish monolinguals, suggesting that L2 learning heightens bilinguals’ awareness of cross-language phonetic differences. The study further revealed that great majority of the students, even the highly proficient ones in English, identified English /æ/ with their Spanish /a/. L2 learners, in making interlingual identifications, perceive TL linguistic features from their NL features, especially when the sounds are perceptually similar; however, when the continuum involves a new sound, the experienced L1 speakers more closely resemble the native English speakers than the inexperienced L1 speakers.

In order to understand interlanguage rules, Selinker [1] suggests studying the L2 learners’ utterances produced in their attempt to communicate meaning in the target language in unrehearsed situations. Other scholars, however, hugely disagreed with dependence on learners’ utterances, rather elicitation of the learners’ intuitions of grammaticality of the TL should be explored [36, 37]. Tarone [36] maintains that learners’ utterances are systematically variable in at least two senses. First, the linguistic context of the learner’s use of phonological and syntactic structures as well as the method of elicitation of data from the learner may have a variable effect on the learner’s production of the phonological and syntactic structures, which consequently may affect IL analysis. The debate on what data should be used to study interlanguage is still unresolved in SLA research [37].

3. Classroom implications of interlanguage

Whichever data is considered authentic and reliable in studying the L2 learner’s interlanguage, the aim of such study is to explore the natural rule of language learning to provide a theoretical basis and practical experiences for classroom instruction [5]. Teachers and researchers use interlanguage studies as tools to establish the linguistic or performance level of learners. In essence, what the learners know and what to learn to approximate native users’ norms are left bare. The varied nature of the interlanguage show that the system can be changed by employing different teaching methods and creating linguistic and social contexts which can facilitate language teaching and learning and delist those that inhibit learning and classroom achievements. Research into the learner’s interim dialect of the TL helps the teacher to effect and affect language teaching tasks through improvements in teachers’ attitudes towards errors and the quality of language teaching and research. One important implication of interlanguage studies is the fact that errors are not misnomer; they are natural part of learning and “red flags” that provide evidence of the learner’s knowledge of the second language [38]. Teachers, who accept errors as a necessary feature of experimentation with the language, provide conducive classroom platforms for more meaningful exchange of ideas. This attitude, according to Firth [29] is showing some results in the work of TESL theorists and researchers – the results which benefit language teaching profession. The study of interlanguage has greatly affected error analysis, for example, Burt & Kiparsky [39] and lately Qurrata’Ain & Widodo [40] reveal that errors within a constituent or a clause affect comprehension of a sentence far less than errors that are made in a constituent or across clause boundaries. The examples below by [39] show how intended relations can be confused by reversing the order of subject and object:

*English language use much people.

The English language uses many people.

Much people use English language.

Researchers have suggested productive ways teachers can deal with errors in students’ interlanguage [12, 41]. First, teachers should not demand grammatical accuracy, particularly at the early stages of second language learning, such demand conflicts with the free use of the target language. Again, the kind of corrections and circumstances under which such corrections are made should be particularly during conversation, as such, contexts must prove informal and friendly than under stringent formal conditions. Over-correcting students when errors are found in their utterances leave them with communication apprehension characterized by exaggerated worry and tension which affect their overall achievements in the classroom. Song [5] observes that active learners employ hypothesis testing to gain recognition of the target language while passive learners are more likely to refrain from learning. The teacher, therefore, should be determined to walk out passive learners through active learners by adjusting their teaching methods to suit learner’s emotion, attitude, and personality. The learner must discover a mismatch between his output and the target language forms at this level of language learning. He can only achieve this if the target language forms are passed through an L1 filter [8]. The learner cannot develop what MacWhinney [7] calls L2 associative network if he is unable to detect a mismatch between his output and the target language. The language teachers, as a matter of urgency, must encourage and guide the L2 learner to achieve this level of language development through computer-enhanced speech processing [42].

The L2 learner’s unsuccessful development of auto-associative network leads to increased perceptible errors in the interlanguage. The developmental nature of the interlanguage reveals degrees of conformity and imprecision of the L2’s hypothesis testing of the target language. Accurate hypothesis testing leads to internalization of the L2 rules. Successful hypotheses become mental constructions that correspond to the rules of the target language while unsuccessful hypotheses, according to [25] are revised or discarded. The implication of this claim, therefore, is that L2 learners are constantly located on an interlanguage (IL) continuum between their L1 and the target language. Highly successful L2 learners achieve a commensurate level of competence while less successful learners become fossilized somewhere along the IL continuum [26].

4. Fossilization

Central to the discussion on the concept of interlanguage is the term, fossilization. First introduced by Selinker [1] fossilization refers to “the long-term persistence of non-target-like structures in the interlanguage of the non-native speakers” ([43], p. 197). Fossilizations are also described as those features which, “though absent from the speech of learners under normal conditions, tend to reappear in their performances when they are forced to deal with different material when either anxious or in an extremely relaxed state Selinker [1] cited in Firth [29]. The non-target-like structures present in form of errors that remain in the learner’s speech or writing over time, no matter what the input or what the learner does. When errors are permanently found in the IL of an L2 learner in a form that is deviant from the TL norms, we say the learner is fossilized in that level of language study. This kind of regression, according to Firth [29] is seen as systematic, and Selinker [1] uses it as evidence for the psychological reality of fossilization and interlanguage. In addition to the term fossilization, Selinker & Lakshmanan [43] further introduced another concept, stabilization which presumes fossilization. The distinguishing feature between the two concepts as introduced by Selinker and his co-researchers is defined in terms of permanence. Stabilized errors are not permanent; rather they are, at a given level of IL development, maintained in L2 learner’s production. At this level of IL development, both the teacher and the learner play significant roles to ensure stabilized errors disappear since stabilization is a prolusion to fossilization. Fossilized errors, however, are permanently established in the IL of an L2 learner [44] in deviant forms from the TL norms and according to [12] persist in his performance irrespective of TL exposure. Firth [29] suggests that the persistence of fossilized forms could well indicate that the interlanguage of a learner who displays this kind of regression is still limited to relatively simple acts of communication. Development of the interlanguage which would facilitate the use of more complex syntactic structures will not take place until its function is extended to such integrative and expressive uses as an affirmation of social identity and expression of psychological need. Fossilization presents at all levels of language study, but records have shown evidence of unceasing occurrences at the phonological level [1, 45].

Chidi-Onwuta & Oluikpe [12] analyzed a sample of the interlanguage of Igbo-English bilinguals to determine manifestation of fossilized errors induced by the target sounds /ʌ, ɵ, and æ/ using data from a passage and a model containing the target sounds given to ten randomly selected staff of the Federal Medical Centre, Umuahia, Nigeria who had their training in English language. Data were analyzed using Spearman’s correlation coefficient statistical method, and mean value of the correct and fossilized errors. The findings revealed a high mean value of fossilized errors for the target sounds which were lacking in the participants’ native language (NL) phonemic inventory /ʌ and ɵ/ with a low mean value of fossilized errors for the target sounds present in the participants’ NL phonemic inventory /æ/. The study suggests that linguistic items present in the target language but absent in the learners’ NL are probable mistake causers which, if not got rid of, with positive input before the learners attain plateaus in IL, would possibly become fossilized items. The pedagogic implication of the study, therefore, is that fossilization is induced by cross-linguistic influence especially if adequate strategies and approaches are not adopted to facilitate learning in order to develop language competence strategies among L2 learners. Other causal factors of fossilization as [12] outlined include lack of corrective feedback, L1 interference, adequate and impoverished quantity and quality of the target language input, method of instruction, (too much based on the NL), L2 learners’ communicative strategies such as avoidance and simplification, inattention to improving L2 proficiency, among others.

Following Chidi-Onwuta & Oluikpe [12, 45] study is Yusuf [46] who studied the effects of some manifestations of interlanguage in English language learning among level 11 students of B.A English at Bayero University, Kano (BUK) and Northwest University, Kano, Nigeria with the aim of using data to reaffirm or refute some of Selinker’s [1] conceptions on interlanguage. Using purposive sampling, they selected five sentences each showing manifestations of one of the five factors or processes of interlanguage. After presentation and analysis, Yusuf [46] found out that learners, in the process of learning English as a second language, manifest all the processes of interlanguage identified by Selinker [1] viz.: transfer of training, mother tongue interference, overgeneralization of the target language rules and strategies of learning. Contrary to Selinker [1] and Tarone [36] claim on transfer of training, learners do not use a word from their mother tongue as a mediator in learning a slightly similar sounding word, rather, they appeal to an already learned second language word. The implication of Yusuf [46] study, therefore, is that when negative input is transferred through any wrong pedagogy, it consequently affects learned material which could inhibit learning of a new item.

Research has shown that fossilization, amidst its persistence in the IL of the L2 learner, is a reversible condition if “the learner is given the necessary positive feedback meant to encourage further attempts at communication” [26 , p. 218]. Following Brown [26] is Demirazen [47] who demonstrated through his study with Turkish English bilinguals using controlled guided pronunciation practice called the audio-articulation method that it is possible to rehabilitate fossilized pronunciation errors brought about cross-linguistic influence. Another significant study is [24] whose studies with Persian EFL IL to establish which particular errors present more: developmental versus fossilized errors, suggests that out of the fifteen prepositional errors identified in the subjects’ compositions, two only were identified as developmental while thirteen were fossilized errors. It implies, therefore, that fossilization is a common feature and an obstacle in second language learning and teachers must guide learners to overcome it. Kayaoglu [48] later proved through his study with Turkish students that learners who de-fossilized could not sustain their newly acquired pronunciation habits, possibly due to their ages. To buttress this suggestion, research has shown that once learners pass the critical period, their pronunciation becomes fossilized. It is invulnerable therefore to claim that permanent defosilization is age dependent [13].

5. The revised interlanguage hypothesis

Discussion on interlanguage hypothesis has evolved across these years in certain ways. It is assumed to impact the field of second language acquisition and was restricted to processes of adult second language learning and learner’s interim dialect of the TL. Today, the discourse has shifted emphasis from adult second language learners to children language acquirers or learners. Tarone [49] cited recent remarkable studies in French immersion programs in Canada which revealed that children produced interlanguages that were apparently fossilized and had substantial influence from their first language. We must recall that exponents of L2 Direct UG Access [16, 17] emphasized that L2 learners can build L2 grammar from the scratch in exactly the same way L1 users build grammars of their native languages from the scratch. It implies, therefore, that language development both in adults and in children learning their mother tongue proceeds in a stage-wise manner. Lately, Jordan & Bittner [50] study on the early interlanguage systems of children learning Dutch and German provides evidence that child learner systems are coherent lexical systems which reveal developmental progress as found in the L2 IL system and driven by the acquisition of the formal linguistic properties of the TL system. The works of Jordan & Bittner [50] and Tarone [49] are responses to Swan [51] linguistic calls that IL hypothesis should be extended from adult second language acquisition to child language acquisition settings where the major sociolinguistic variable is the absence of peers who are native speakers of the TL. Their study revealed that cognitive processes like the strategies of language transfer, simplification, and overgeneralization of TL rules affect the surface forms of the second language speech of children in their program. The social implication of the study is that ILs will develop as dialects in their own right. In Ref. [51] shows that IL system is not peculiar to adult learners, rather children who pass through the same process of acquiring a second language share same cognitive processes that shape adult IL system. The study supports claims about systematic and permeable nature of ILs.

Swan [51] who studied IL of children as they attempted to communicate meaning in French revealed engagement of several learning strategies (language transfer, overgeneralization of TL rules, and simplification). It is in the consistent use of these learning strategies that [51] found their IL as systematic as adult IL. They also observed that in some cases, several learning strategies were operated simultaneously or sequentially, resulting in greater IL stability. Selinker & Swain called for future investigation into the degree to which sociolinguistic settings can affect change over time in children’s IL system. Their findings suggest that the sociolinguistic context (free from anxiety and without native speakers) with which the children communicated among themselves induced stability of the IL.

Earlier, Ellis [52] examined the interlanguage grammar of Kindergarten students of Mondial Education Semarang to establish the errors that usually occur in their speech and to describe the factors of the interlanguage grammar that influence their English. His findings revealed that Bahasa Indonesia interferes with the target language in all levels of language study. The structure of Bahasa Indonesia inhibited their performance in English. The social environment of the children (Indonesia) induced negative evidence in the TL, supporting a correlation between the degree of fossilization and the degree of L1 influence through the production of the target language forms.

The works above appear to establish cross-linguistic influence and inadequate knowledge of L2 in children’s IL but following the argument about a biologically determined period of life when language can be acquired more easily and beyond which time language is increasingly difficult to acquire [13] children have increased tendency to achieve permanent defossilization. Again, a deep look into the social nature of children’s IL reveals that they are not given enough opportunity and incentive to produce what [53] calls “comprehensible output” –attempts to produce interlanguage in meaningful communication with others.

In our review of the concept of interlanguage, we must note that UG does not produce IL as it does natural languages, IL therefore must be produced by cognitive processes other than UG. Tarone [49], however, sees IL from a different perspective. He remarks, “as natural languages, interlanguage has to obey language universals.” Debate on whether interlanguages are products of the same UG that produce native languages is still ongoing, and we are not inclined in this study to delve into the controversy now.

A very significant shift in IL discussion is seen in a growing interest in the way IL forms change in different social contexts [54]. Tarone [54] observes that a more fluent, grammatical, and transfer–free IL is evidenced in some social contexts than in others. Variationist researchers, according to Tarone [54], have argued that because production of given interlanguage forms varies systematically in relation to social contexts, task, topic, interlocutor, etc. Researchers must be ready to document the contexts at play in each elicitation and use adequate statistical tools to model interlanguage.

In conclusion, the study of interlanguage, both in adults and in children is central to second language acquisition and learning. It provides evidence of the learner’s knowledge of the second language as well as information about different linguistic systems aimed at improving classroom experiences, but researchers must ensure usage of adequate data for IL study.