According to Brown in order to analyze a learner language, it is important to understand the distinction between errors and mistakes. A mistake refers to a performance error that is either a random guess or a slip, seeing as a failure to use a known system correctly. Mistakes can happen in both native and second language are the result of some sort of temporary breakdown, hesitations, slip of the tongue, random ungrammaticalities, or imperfection in the process of producing speech and can be recognized and corrected by native speakers. However errors are idiosyncrasies in the language of the learner. They are actually deviation from adult grammar of a native speaker. It shows the competence of the learner. Then mistakes are referred to performance errors in which the learner knows the system but fails to use it while errors are the result of one's systematic competence. It means that the learner's system is incorrect.
According to James cited by Brown (2007) the learner is not able to self correct the errors while mistakes if called, can be corrected by the learner. Then error correction can be used as a means for identification of errors and mistakes. However, Brown noted that making a distinction between errors and mistakes is difficult because if no such error correction occurs, we are not able to distinguish those (2007).
Keshavarz (2008, p.49) also noted that there is a distinction between errors and mistakes. Errors are rule governed, systematic in nature, internally principled and free from arbitrariness. They show learner's underlying knowledge of the target language that is his transitional competence. In contrast to errors, mistakes are random deviations and unrelated to any system. They are related to performance of the learner and might occur in the speech and writing like slip of the tongue, slip of the ear, slip of the pen, and false start. According to Keshavarz (2008) mistakes are due to non-linguistic factors such as fatigue, strong feeling, memory limitations, and lack of concentration and so on. These kinds of mistakes can be corrected by the learner if brought to his attention.
He considered that distinguishing between learner's errors and mistakes has always been problematic for teachers and researchers as Corder (1967) and Brown (1987) cited in keshavarz (2008) maintained this problem too. Nevertheless most error analyst use a general criterion for distinguishing between errors and mistakes" the frequency of occurrence" that is: errors which are low frequent are considered as mistakes or performance errors and those which are high frequent are systematic errors. However this criterion alone is not enough for distinguishing between errors and mistakes because low frequency of certain error may due to the low frequency of grammatical patterns or avoidance strategy that a learner uses. Then error analyst should consider casual factors of learner's deviant structures as a means of distinguishing between errors and mistakes.
According to Corder (1997) cited in Park (2010) errors refer to learners' underlying knowledge of the language and mistakes refer to incorrect forms caused by memory lapses, slips of the tongue and other instances of performance errors. Corder considered that learners can correct their own mistakes but they cannot correct their errors because they do not have enough knowledge to distinguish their own utterance and that of the native speaker. He also pointed out two explanations with regard to learner errors. First, "the occurrence of errors is merely a sign of the present inadequacy of the teaching techniques" (p. 163). That is, if it were possible for teachers to achieve a perfect teaching method, there would be no occurrence of student errors in the target language. The second explanation is that despite teachers' best efforts, the occurrence of errors is inevitable because errors occur for many reasons. The reasons can be: interference from L1, overgeneralization, an incomplete knowledge of the target language, the complexity of the target language, and fossilization. Therefore, teachers should be more concerned with how to deal with students' errors than the simple identification of them. For this study, I will use the terms "errors" and "mistakes" interchangeably because sometimes it is difficult to distinguish students' errors from mistakes.
Identification of Errors: (Categories of Error)
Corder (1971) cited by Brown (2007) provided a model for identification of errors. Based on his model any sentences uttered by the learner can be analyzed for idiosyncrasies. According to his model there are two types of error: overt and covert errors. Overtly incorrect utterances are ungrammatical at the sentence level and covertly incorrect utterances are grammatically correct at the sentence level but are not interpretable within the context of communication. Then overt errors are sentence level and covert errors are discourse level errors.
Burt (1975) cited in Park (2010) categorized errors as global and local ones. Global errors are errors that hinder communication and affect the overall organization of the sentence such as wrong word order, missing, wrong or misplaced sentence connectors however local errors do not usually hinder communication and affect single elements in a sentence such as error in noun and verb inflections, articles, and auxiliaries. He mentioned that the correction of global error clarifies the intended message more than the correction of several local errors. Moreover he argued that teachers should correct high frequent errors first. From another perspective Chaudron (1977, p.32) cited by Park (2010) classified the range of errors from" linguistic (phonological, morphological, syntactic) to subject matter content (factual and conceptual knowledge) and lexical items". Valdman( 1975) cited in (salim shahin) provided the same division as Burt and said that in the case of global error the communication between the student and teacher will be blocked and the student requires correction but in the case of local error, communication between the teacher and the student will not be blocked and it is up to the teacher to require the correction of the error, or let the error passes.
Hammerley( 1991) cited in (Salim Shahin, 2003) classified errors into surface and deep. According to him surface errors need minor corrections and explanations while deep errors require explanation of why the error was made and what the correct form is. Hendrickson (1978) cited in (Salim Shahin, 2003) divided errors into three main types. 1) errors that hinders communication; 2) errors that have highly stigmatizing effect on the listener or reader but do not hinder communication; 3) errors that can be described as lapses that students usually have in their utterances. These errors are quite common in the speaker's utterances yet they hardly block communication between the speaker and his interlocutor.
Mackey et al. (2000) cited by Park (2010) categorized four types of errors in their analysis of L2 interactional data as phonological, morphosyntactic, lexical, and semantic ones.1) Phonological errors were non-target like pronunciation; 2) morphosyntactic errors were omitted plural-s and the preposition in; 3) lexical errors were inappropriate lexical items; 4) semantic errors were incorrect meaning or expression.
Prabhu (1987) cited in (Salim shahin, 2003) divides errors on the basis of their treatment rather than their nature. According to his division there are two kinds of error: systematic and incidental errors. Systematic errors are those that deviate from the native speaker's form and the teacher uses linguistic explanation, long interruptions, and exemplification to help and correct learner's error. But Incidental errors do not require linguistic explanation or exemplification from the teacher, such as when the teacher corrects a pronunciation error or when he raises his eyebrows to draw the student's attention to the error.
Moreover Lennon (1991) cited by Brown (2007) identified different categories for description of errors:
Errors of addition, omission, substitution, ordering
Level of errors: phonology or orthography, lexicon, grammar and discourse
Global and local errors: an error that hinders communication or prevents hearer or reader from understanding some aspect of a message is global and an error that does not prevent a message from being understood, usually due to a minor violation of one segment of a sentence, allowing a hearer or a reader to make an accurate guess about the intended meaning is a local error. (burt&kiparsky, 1972 cited by Brown, 2007)
Domain and extent error: domain error is the rank of linguistic unit from phoneme to discourse that must be taken as context in order for the error to become apparent. Extent error is the rank of linguistic unit that would have to be deleted, replaced, supplied, or recorded in order to repair the sentence.
Other educators like Allwright (1975) cited in (Salim Shahin, 2003) believed that errors should be treated on the basis of their frequency, rather than on their classification. Then errors of high frequency should be given more attention and emphasis than errors of low frequency.
Some Definitions in Error Correction:
There are different terms when providing feedback in response to learner error in second language acquisition. These are repair, treatment, feedback, negative evidence and correction. Brown (2007, p.388) defined repair as correction by the learner of an ill-formed utterance, either through self-initiated repair or in response to feedback. Chaudron (1988) cited in Dabaghi (2006) defined treatment as any teacher behavior that follows error and shows the error to the learner. Ellis (1994a) cited by (Dabaghi, 2006) also identified feedback as a general cover term in which listeners provide information on the reception and the comprehension of the messages.
Lightbown and Spadal (1999) cited in Dabaghi (2006) defined corrective feedback as an indication to the learners that they use target language incorrectly. According to Ellis (1994a) cited by Dabaghi (2006) correction has a narrower meaning than these terms. It is any attempt to provide negative evidence in the form of feedback that draw learner's attention to the errors they have made.
As Schachter (1991) cited by Dabaghi (2006) said corrective feedback, negative evidence and negative feedback are terms that are respectively used in the field of language teaching, language acquisition, and cognitive psychology. Most of the researchers use these terms interchangeably. In the present research we use error correction and corrective feedback interchangeably whenever the general sense of feedback provision was involve.
Types of spoken error correction:
According to Lyster & Ranta's model (1997) cited by Coskum (2010) the types of spoken error correction are as follows:
Explicit correction: clearly indicating that the student's utterance was incorrect, the teacher provides the correct form.
S: there is a little milk in fridge.
T: + in the fridge
Recast: the teacher implicitly reformulates the student's error, or provides the correction without directly pointing out that the students' utterance was incorrect. In other words it is the teacher's reformulation of all or part of a student's utterance, minus the error.
S: he like pop-music.
T: yes, he likes pop-music
Clarification request: the teacher indicates that the message has not been understood or that the student's utterance included some kind of mistake and that a repetition or a reformulation is required by using phrases like "Excuse me?"
S: there aren't many /hotÄ±ls/ in this town.
T: again? ô€‚‘
Metalinguistic clues: the teacher poses questions like "Do we say it like that?" or provides comments or information related to the formation of the student's utterance without providing the correct form. Metalinguistic clues contain comments, information or questions related to the well formedness of the student's utterance, without explicitly providing the correct form. Metalinguistic comments generally indicate that there is an error somewhere. Metalinguistic information generally provides either some grammatical metalanguage that refers to the nature of the error or a word definition in the case of lexical errors. Metalinguistic questions also point to the nature of the error but attempt to elicit the information from the student.
S: there isn't any books.
T: + there isn't görünce uncountable, yani sayÄ±lamayan bir ÅŸey kullanmamÄ±z gerekiyormuÅŸ.
Ds: there isn't any money
Elicitation: the teacher directly elicits the correct form from the student by asking questions (e.g. "How do I ask somebody to clean the board?"), by pausing to allow the student to complete the teacher's utterance ("e.g. He is a good â€¦..)" or by asking students to reformulate the utterance (e.g. "Can you say that again?"). Elicitation questions differ from questions that are defined as metalinguistic clues in that they require more than yes/no response. It refers to at least techniques that teachers use to directly elicit the correct form from the students. First teachers elicit completion of their own utterance by strategically pausing to allow students to "fill in the blank" as it were. Second, teachers use questions to elicit correct forms. Such questions exclude the use of yes/no questions. The question likes Do we say that in English is metalinguistic feedback, not elicitation. Third teachers occasionally ask students to reformulate their utterance.
S: there are a few books in my /lÄ±brari/
T: in myâ€¦? ô€‚‘
Repetition: the teacher repeats the student's error and changes intonation to draw student's attention to it.
S: How much money do you have in your /pakÄ±t/?
T: /pakÄ±t/? ô€‚‘
Sources of Error:
According to Brown (2007, p.263) there are four sources of error:
1) Interlingual transfer: it is a significant source of error for all learners. It is the effect of one language (usually the first) on another (usually the second). At early stages of learning a second language, the linguistic system of the native language is the only system that a learner is familiar with. Then the native linguistic system can interfere or transfer and cause some errors in second language.
2) Intralingual transfer: Another source of error which goes beyond the interlingual transfer is intralingual transfer. It is the effect of forms of one language (usually the target language) on other forms within the same language. According to Odlin (2003), Jaszczolt (1995) and Taylor (1975) cited by Brown (2007) once learners have begun to learn the system of second language, it is the system of second language that cause error. In fact intralingual transfer that is generalization within the target language occurs.
3) Context of learning: The third source of error is context of learning which is called false concept by Richard (1971) and induced error by Stenson (1979) cited by Brown (2007). It overlaps two types of transfer and refers to classroom, materials and social situation that can lead learners to make faulty hypotheses about the language.
4) Communication strategies: The forth source of error is communication strategies. They are production strategies that learners use to enhance getting their message across, but these strategies can become a source of error.
Moreover Corder (1975) cited by Keshavarz (2008, p101) distinguished three types of sources of errors:
Inter lingual errors which are caused by first language interference
Intralingual errors that are caused by the learner's generalizing and over generalizing particular rules.
Errors caused by faulty teaching techniques.
Another classification was considered by Dualy and Burt cited by Keshavarz (2008, p. 101) for second language learners' error which are called goofs:
Interference-like goofs: errors which show native language structure and are not found in first language acquisition data of the target language.
L1-developmental goofs: errors that do not reflect native language structure but are found in L1 acquisition data of the target language
Ambiguous goofs: those errors that can be as either interference-like goofs or L1 developmental goof.
Unique goofs: errors that do not reflect L1 structure and also not found in L1 acquisition data of the target language.
Significance of errors:
Many scholars in the field of error analysis have stressed the significance of second language learners' error. For example Corder (1967) cited by Keshavarz (2008) remarked that errors are significant in three ways: First errors are significant to the teacher in that they help him to understand how much the learner has progressed and what remains for him to learn. Second errors provide evidence for the learner to recognize how language is learnt or acquired and what strategies and procedures are used by the learner to discover the language. Third errors are important to the learner himself in which they are a device the learner uses in order to learn.
Moreover Richard (1971) cited by Keshavarz (2008, p.45) noted that errors are significant and of interest to: 1) Linguistics, because according to Chomsky the study of human language is the best way of understanding the human intelligence; 2) Psychologists, because by comparing children's and adult's speech, they can examine the nature of the mental processes that seem to be involved in language; 3) Teachers, because by analyzing learner's errors, they would be able to discover their difficulties and devise a method for comparing them.
Besides, Jain (1974) cited by Keshavarz (2008) maintained 2 reasons for the significance of error:
Understanding the process of second language acquisition
Planning courses incorporating the psychology of second language learning.
Other researchers like Dulay and Burt (1975) cited by Dabaghi (2006) stressed the significance of error by indicating two major reasons: 1) error provides data from which interferences about the nature of the language learning processes can be made, and 2) it shows to the teachers and curriculum developers in which part of the language, learners have difficulty with and which error types detract most from the learner's ability to communicate effectively.
Historical perspectives of error correction:
Russel (2009) noted that error correction is a controversial issue in the field of second language education (SLE) and second language teacher education (SLTE). How to correct errors depends on the methodological perspectives of teachers.
In behaviorist teaching models such as audio lingual method that was popular in 1950s and 1960s, error correction was stressed at all cost. Behaviorists believed that errors were inevitable but they tried to provide the correct form immediately. Brooks (1960, p.56) cited by Russull (2009) considered that "like sin, error is to be avoided and its influence overcomesâ€¦ the best way to overcoming errors is to shorten the time lapse between the incorrect response and a presentation of correct form". However in 1970s the value of grammar instruction and error correction in behaviorist model was questioned. Russull (2009) considered that in 1970s and 1980s some scholars claimed that error correction was not only unnecessary but also harmful to second language acquisition. The most well known model which was against error correction was Stephan Krashen's monitor model which has five hypotheses about language learning. Rashtchi & keshavarz (2007) noted that the learner's affective and emotional status can act as filters which do not allow easy absorption of input. It means that affective filter will hinder the process of learning (p.76). Moreover they noted that according to natural order hypothesis all linguistic elements and skills are learned in a predictable order and this order is not influenced by the native language of the learners.
Russell (2009) mentioned that Terrell created the natural approach which is a method that emphasizes the development of communicative competence. According to his approach affective rather than cognitive factors are primary concern in the language classroom and correction of learners' error is negative in terms of motivation, attitude and embarrassment. In this approach teacher never correct the learner oral errors. Then learners themselves should study grammatical structures in order to correct their mistakes.
Communicative language teaching approach became popular in 1980s and like natural approach focuses on communicative competence and notional functional concepts over the instruction of grammatical structures. (Richards &Rogers (1986) cited by Russell (2009). The objective in CLT is the development of fluency and acceptable language use and since emphasize is on meaning over form, error correction is not of primary importance. However when learners' accuracy is assessed, it is always done in context. (Omaggio Hadley, 2001 cited by Russell), and in 1990s some researchers assert that explicit grammar instruction, error correction and focus on form could promote SLA. (Aljaafreh & Lantolf, 1994; Doughy & Varela, 1993; Ellis, 1993, 1994; Fotos, 1994; Long 1996; Schmidt, 1990, 1993, 1995; Sharwood Smith, 1993).
Error correction and second language acquisition:
According to Dabaghi (2006, p. 25) there are different views and opinions about language teaching and learning in which the changes in methodologies, materials and attitudes toward second language learning have always been important to language experts. Since error correction cannot be separated from these views and opinions, it must be discussed in terms of its relation to the theories of second language acquisition. In the following sections, we can see a number of second language acquisition theories and models that in some way affect our understanding of error correction.
Contrastive analysis model:
In the era of contrastive analysis and audiolingualism, there was a negative approach towards errors. According to Stern (1983) cited in Dabaghi (2006, p.25) some of the scholars during 1950s and 1960s had a puritanical perspective about errors committed by second language learners and as Brooks (1960) cited by (Dabaghi, 2006, p.25) said "like sin, error is to avoid and its influence overcome but its presence is to be expected.
According to Dulay, Burt, and Krashen (1982) cited by Dabaghi (2006), in contrastive analysis, there is a comparison between learner's native and target language and these differences cause the majority of an second language learners errors. Based on behaviorist view which was prevalent at that time, learning is a process of habit formation and errors from first language habits interfere with the learner's attempt to learn new linguistic behaviors. Then to avoid errors, teachers should prevent and correct errors and shorten the time lapse between the incorrect response and the correct model. But as Dabaghi (2006, p.26) said, paying too much attention to prevention of errors and teaching learners to use language creatively in response to expected and unexpected stimuli in the environment are the drawback of contrastive analysis model and structural approach and caused the very foundation of contrastive analysis as a result of Chomsky's transformational generative grammar, which emphasized the active participation of the learner's mind in processing data. In short, according to CA, learning is a process of habit formation and interfering language habits with learner's attempts to learn new linguistic behaviors cause errors. Then the audio lingual approaches to teaching aimed to prevent learners from committing errors and believed that errors should be corrected immediately.
Inter language model:
In inter language era the second language learner was considered to be an autonomous creator of a language system who has a built-in-syllabus (Corder (1981) cited by Dabaghi (2006, p. 26)). It means that a learner is equipped with an internally programmed sequence which is sometimes in accordance with what teacher teaches and sometimes contradictory and learner follows his/her own build in syllabus. According to this view errors are the evidence of the learner's present transitional competence and manifest the way the learner processes the input in her linguistic environment.
Dabaghi (2006, p.27) mentioned that the inter language view of language acquisition believed on the whole role of language learning to the language learner and weaken the impact of negative evidence in second language development. According to this view, if negative evidence is provided within the syllabus predetermined in the mind of language learners, it can be effective otherwise it causes frustration and confusion for the learner and teachers. This view was strengthened by non-interventionist view taken by Dulay and Burt (1973), Krashen (1983) and Prabhu(1987) cited by Dabaghi (2006) who argued that grammar instruction should be abandoned in order to let the learner acquire the language from untutored language setting. The teacher should only provide the learner with opportunities for natural use of language. To sum up, the inter language model considered second language learners as self governing creator of a language system who follow their own built in learning program which can sometimes benefit from error correction and sometimes not.
Krashen's Input Hypotheses:
it is commonly known that for the acquisition of a second language, input that is provided either by a teacher or by another learner is essential. Corder (1967) cited by (Faqeih, 2012) distinguished input from intake. According to him, what is available to the learner is called input and what is actually internalized is called intake. According to Faqeih (2012) Krashen proposed the most influential theory of the role of input in second language acquisition in 1980. He noted that for the acquisition of second language, learners must be exposed to comprehensible meaningful input which contains linguistic data that are a little beyond learner's knowledge (i+1) where i is the second language learners current linguistic competence and (i+1) is the next level of that competence achieved with comprehensible input ( Krashen, 1985 cited by Faqeih, 2012). He also suggested that production of the target does not directly aid acquisition. Krashen considered that learners can make use of three kinds of contextual information: extra-linguistic information that is learners' knowledge of the world and previously acquired linguistic competence; the input that can be available via interaction; and interaction in which meaning has to be negotiated e.g. when there is a communication problem.
Krashen (1985) cited by (Faqeih, 2012) also distinguished "learning" and "acquisition". He considered that acquisition uses unconscious processes and those grammatical rules are not helpful. In other words, second language is acquired more like first language and parents focus on communication and meaning instead of focusing on explicit instruction of the language. He also implied that if input is understood and there is enough of it, the necessary grammar is automatically learnt. He also argued that error correction (= negative evidence) do not improve second language performance. On the other hand "learning" is the conscious process that involves the memorization of many formal grammatical rules and error correction can have a role in this. According to him learning leads to grammatical and mechanical knowledge of the language, but it does not lead to fluency. Note that although Krashen distinguished between learning and acquisition, but this thesis uses them interchangeably.
According to Faqeih (2012) Krashen's input hypothesis is supported by some scholar such as Shwartz (1993) and Truscott (1996). They noted that for learners to acquire second language, only positive evidence that is a model from the target language is sufficient and negative feedback such as implicit and explicit feedback to any non target like feature in learner's speech does not help learning and that structures learned through error correction cannot become part of internal grammar. They also proposed the negative effects of error correction such as confusing the learners, causing over use of a particular form or interfering with natural language acquisition processes. Others like Lightbown & Spada 1993 cited by Faqeih (2012) said that excessive use of error correction lead to lower motivation. According to Krashen's affective filter, classroom teachers should not focus intentionally or explicitly on errors of language from during class but should instead provide comprehensible input to learners.
Other scholars such as Mitchell & Myles (2004) cited by Faqeih (2012) criticized this hypothesis because of lacking in empirical evidence and untreatable owing to its vagueness. Moreover White (1987) criticized Krashen's input hypothesis for not considering the possible benefits of providing rules, said that certain types of errors may need rules instruction. Others like Ellis, Tanaka, & Yamazaki (1994) cited by Faqeih 2012 considered that although interaction can solve communication problems through negotiation and increased comprehension, it does not mean that increased comprehension automatically leads to L2 acquisition; that is, learners may not necessarily retain the comprehended target language.
Long's Interaction Hypothesis:
Based on Krashen's input Hypothesis in which input is important for acquisition to take place, Long suggested his own interaction hypothesis (Long, 1996, 2007 cited by Faqeih (2012)). He considered that input is not sufficient on its own for language learning to occur and some type of negative evidence may be beneficial role in acquisition and more attention should be given to the interaction that learners are engaged in. In 1996 he updated his hypothesis and mentioned that some controversial strategies such as repetitions, confirmation checks, comprehension checks, clarification request or recasts can solve communication problems. This kind of interaction involved components like negotiation, recasts, and feedback. According to Long (2007) cited by Faqeih (2012) interactional feedback is very important aspect for language development because it helps learners become aware of their error, and notice the mismatches between their inter language and the target language. It also encourages learners to hypothesize the correct form and test them and modify their inter language. Long (2007) cited by Faqeih (2012, p.29) claimed that for error correction to be affective, simultaneous focus on form and meaning should be provided in a classroom context and the best strategy for the negotiation for meaning is recast, Because it is implicit and does not interrupt the flow of interaction.
Dabaghi ( 2006, p.28) mentioned that according to Long, when there is a conversation between two interlocutors ( native speakers and non native speaker) there are both similarities and differences in the form of grammatical complexity between them. They use some conversational tactics such as clarification request, repetition or comprehension checks. It shows that the non native speaker or less component interlocutor is experiencing comprehension problems and using these tactics is very useful in language learning. In the process of interactional adjustments, both interlocutors make effort to understand each other. It means that they adjust their input to make it more comprehensible that is i+1 in Krashen's term. However this hypothesis was criticized by Braidi (1995) cited in Dabaghi (2006, p.28) because of too much attention on analyzing the meaning aspect of interaction between native speakers and non native speakers and little attention to grammatical aspects of learner's language. Then he reformulated the hypothesis and paid more attention on features that link input and environment with learner cognitive factors and recognized the role of negative evidence in learning and introduced the notion of selective attention to explain how input becomes intake. To sum up, negative feedback directed at comprehensible input during negotiation work may lead to second language development for certain structure features.
Long interaction hypothesis has received theoretical and empirical support. On the theoretical grand, Doughty (2001) cited by Faqeih (2012) supported the use of recast which give learners the opportunities to engage in form-meaning mapping. Carrol (2001) cited by Faqeih (2012) also supported updated version of this hypothesis. Moreover on the empirical evidence Mackey (1999) cited by Faqeih (2012) in her investigation of question formation showed that interlocutors who are involved in directed negotiation were able to develop their second language knowledge of question formation faster than non-interactors.
Sheen (2006) cited by Faqeih (2012) criticized Long interaction hypothesis in which Long claimed that for effective error correction, learners require to be equally attending to form and meaning, because according to her it is often not clear whether the feedback is a result of communication breakdown or it is teacher's choice. She also noted that teacher's choice for the error correction is much more common than conversational feedback. There is some empirical evidence against effectiveness of implicit feedback. For example Ellis (2006) and Sheen (2006) cited by Faqeih (2012) suggested that implicit error correction does not promote acquisition. Moreover Long argued that explicit error correction interrupt the flow of communication and does not help learning. Moreover Sheen (2006) and Ellis (2006) suggested that metalinguistic feedback by the teacher do not interfere with the communicative flow of the activity in the classroom.
Mclaughlin's information processing model:
According to Mclaughlin (1987) cited by Dabaghi (2006, p.29) information processing may be either controlled or automatic. As noted in Longman teaching language and applied linguistics (p.257) in this model information and meaning are stored, organized and retrieved from memory and different decoding will take place during reading and listening. Based on this view, learning involves a shift from controlled towards automatic processing which results in restructuring of the linguistic system of the second language learner. In sum, it seems that error correction handled by control processing and later as a result of practice, it becomes automatic and part of learner's inter language.
Anderson's Declarative and procedural knowledge:
Error correction can also been discussed through declarative and procedural knowledge (Dabaghi, 2006, p. 29) these knowledge are related to controlled and automatic process. Anderson (1983) cited in Dabaghi (2006, p.29) considered declarative knowledge as a knowledge that refers to learner's information about a form and it has neither been automatized nor integrated into his /her inter language system. Procedural knowledge is the knowledge that has been auomatized and made readily available for use as implicit knowledge. As a result of practice the declarative becomes proceduralized and available for unconscious use. He believes that there are 3 kinds of memory for these knowledge and they are stored in different ways. In summary, correction based on this view, provides the learner with the knowledge about the corrected form and later, as a result of practice, this knowledge will be integrated into learner's inter language.
Schmidt's noticing hypothesis:
According to ( Faqeih, 2012 )" given that regular interaction as in Long's (1996) interaction hypothesis depends on works through learner-internal factors such as noticing, empirical research investigated the relationship between noticing and learning in theses contexts".
Schmidt (1990, 1995) cited by Faqeih (2012) found this relationship and argued that learners need to pay attention to the gap between input and what they produce. Then the process of converting input into intake was the basis of Schmidt's noticing hypothesis. Following his model, interactional feedback became important since it helps direct the learner's attention towards the mismatch between the target input and their own interlanguage form. Based on his experience, Schmidt as an American learner of Portuguese in Brazil investigated that instruction, interaction, and correction influenced his learning of Portuguese and during interaction with native speaker, the target features in the input were almost acquired. Other SLA researchers also found the relationship between noticing and L2 development in the presence of interactional feedback. For example Mackey (2006) cited by Faqeih (2012) considered that, in the presence of interactional on problematic L2 forms, learners pay attention those form more than when feedback is not provided, but there was no clear indication that learning follows noticing for some learners' reported noticing but did not develop and a few learners in the control group developed but did not report noticing the target items. It was a warning to Schmidt's hypothesis regarding the effectiveness of noticing on learning the target items. Then he claimed that "some forms may not be noticed until learners are developmentally ready, and noticing in the input could be affected by different factors: instruction, frequency, perceptual salience, skill level, task demands and comparison.
Schmidt and Frota's (1986) empirical research cited by Faqeih (2012) supported the effectiveness of noticing hypothesis in L2 acquisition. In this research based on Schmidt's experience, producing L2 grammatical forms was not directly related to the grammar that he received as input, but the grammar that he had noticed.
from the theoretical point of view Gass(1988) considered noticing as the first stage of language acquisition; Bostone (1994) as the "gateway to subsequent learning" (p.100); Lynch(2001) as an important component of successful language learning and Vanpatten'(2007) input processing relies on the basic notion of attention to forms to be learnt (cited by Faqeih (2012)). However there is some criticism to this hypothesis. For example Truscott (1998) cited by Faqeih (2012) mentioned that "the foundation of the hypothesis in cognitive psychology are week and is not based on any rational theory of languages" (p.104) and this hypothesis is too vague to determine what learners must notice. He said that noticing is necessary for metalinguistic knowledge but not language competence and more investigation is needed for understanding noticing in SLA.
Swain's output hypothesis:
According to Faqeih (2012) Swain (1985, 1995, 2000, 2005) considered the importance of output hypothesis based on both formal and informal observations in the context of immersion programs in Canada and noted that language will be developed if we push learners to produced output. Moreover it helps learners to pay attention to the gap between their interlanguage and their target language and test their interlanguage hypothesis. She also claimed that language production forces learners to move from the semantic strategic to syntactic use of language as a result of the three functions of output. She also considered the importance of using corrective feedback such as clarification request, as these can promote pushed output and thereby help learners to develop their interlanguage.
Teachers' versus students' perception of error correction:
One of the important factors in the research or error correction is how teachers' and learners' perspectives differ regarding error correction. (Russull, 2009, p. 27)
According to Schulz (2001) cited by Russell (2009) teachers' and learners' different belief system can hinder learning. In her 2001 study among the U.S. and Colombian cultures, she investigated the teachers and learners perceptions about the instruction of grammar and oral error correction. She found that Colombian learners expect more grammatical instruction and focus of form activities in their foreign language classroom than their American counterparts. Moreover learners from both cultures believe that the teacher is an expert knower whose role is to explain and provide feedback. They also preferred that teacher correct their both written and oral errors during the class. Based on the research on teachers' perception, she found that there is a mismatch between belief of Colombian and American teachers regarding grammar instruction that is Colombian teachers emphasize the importance of grammar instruction more than American ones. However both Colombian and American teachers believed that written errors should always be corrected. Conversely only about half of the teachers from both cultures preferred oral error correction in class which is against the learners' belief. Based on her research Schulz cited by Russell noted that learners have specific expectations, beliefs and attitudes and if these are not met, learners' success at learning may be hindered. She claimed that it is the teachers' responsibility to understand learners' belief and expectation about error correction.
Lasagabaster & Sierra (2005) cited by Russell found a mismatch between teachers' and learners' belief about oral error correction. According to this research learners believed that constant correction is not helpful because they think it inhibits language production. However they like to be corrected by their teacher in a selective and explicit way. Students believed that teachers should devote more time to each correction and they should also use more strategies when they want to correct oral errors. Teachers on the other hand were concerned about not correcting the students for fear of including language anxiety. Moreover they believed that it is not both practical and beneficial to correct each and every error that students commit. Their main desire was to achieve a balance between the gravity of the students' error and allowing sufficient learner talk time. According to their research teachers need to provide more explicit and direct error correction to their students specially correction that are noticed and understood by the learners.
Oladejo (1993) attempted to identify the preferences and expectations of intermediate and advanced ESL learners regarding error correction. Then two sets of data were examined in his study. The first set was collected by his student L.k Lim in 1990. He investigated the attitudes and expectations of secondary school pupil in Singapore to error and error correction in English language. The second set of data was collected by James himself. He asked undergraduates of National University of Singapore from five faculties who had registered for proficiency course in English. By means of a questionnaire he concluded that teachers' opinion and classroom practice regarding error correction do not always match the perceived needs and expectations of learners. Such mismatch could contribute to lack of success in language learning. Moreover he noted that our approach to error correction in language classroom cannot afford to be rigid. According to him if error correction is to be effective classroom practice cannot afford to be based rigidly on any standardized practice derived from the opinions of linguists and teachers alone, but it must be flexible enough to incorporate the perspectives and needs of language learners.
Lee (2005) considered that much of the error correction research has done on teachers' strategies and their effects on students' writing and few studies found students' beliefs and attitudes about teachers' feedback on error. Then by means of a questionnaire and follow up interview, she investigated L2 secondary students' perceptions, beliefs, and attitudes about error correction in Hong Kong. She concluded that most students wish their teachers to mark and correct errors for them and believed that error correction was primarily the teachers' responsibility. Moreover the study demonstrated that a huge gap between teachers' practices and student preferences in error correction does not exist.
Diab (2005) considered EFL university students' preferences for error correction, paper making techniques, and their beliefs about what constitutes effective feedback. He collected the data by a questionnaire consisted of 2 parts. The first part was a 12-item background questionnaire in order to gain background information and the second part was the modified version of Leki's questionnaire. His findings supported the general idea about error correction that the L2 learners expect surface-level error correction from their teachers and believed that such feedback is beneficial. He considered that it is teachers' responsibility to be aware of their students' perspectives of what helps their progress and to somehow incorporate theses perspectives in their teaching. Therefore incorporating classroom discussion on error correction, feedback and writing can be essential in helping L2 teachers become familiar with their students' belief about what constitutes effective feedback and modifying or reinforcing these beliefs accordingly. (Diab, 2005)
Incecay and Dollar (2009) investigated the foreign language learners' beliefs about grammar instruction and error correction. For the purpose of their study, they used a questionnaire of Loewen et al (2009). The result showed that learners believe that error correction is useful in language learning process.