Turkish Parents ’ Reaction to their Children ’ s Errors Sedigheh Vahdat

There was a debate regarding first language (L1) acquisition in the area of child error question that parents are not sensitive to their children’s errors. However, some scholars believe that the reverse is true. To address these discrepancies, this study focuses on Turkish parents’ correction of their children’s errors. The data for the present study was collected from 70 parents in one of the kindergartens of Urmia, Iran. These parents were supposed to answer three main questions regarding children’s error correction: 1) What is your strategy in correcting your child errors?, 2) What kinds of errors are more common, grammatical, pronunciation or lexical?, and 3) Does each type of error lead to uptake? The results of the study revealed that parents neglected their own children’s errors and the rate and frequency of phonological errors were more than grammatical and lexical errors. However, uptake was low after parents’ reaction to the erroneous utterance produced by the children.


Introduction 1.
Language is learned in conversation, not in a vacuum.Conventional forms and functions of language are acquired by children through conversing with others.Furthermore, they learn how to express their own intentions and how to interpret other people's intentions through conversation.Children's language acquisition device (LAD) is triggered by adults' speech, which is regarded as a prerequisite environmental factor.It is believed that since the speech children receive from their environment is too poor, i.e. poverty of stimulus, children must have an innate LAD (Birjandi & Nasrollahi, 2012).
There have been various ideas regarding the input received by the child.Nativists, for instance, say that the linguistic information is the only available input for the child.Moreover, nativists claim that positive input and negative evidence are same in the child view (Saxton, 1997).Positive data is described as "any input utterance which models grammatical structures, excluding all instances of negative evidence" (Saxton, 1997).While negative data provides the true adult type directly following a child wrong, positive data provides the right form in all other discussion circumstances.Negative data happens directly conditional on a child wrong, (morpho-syntactic or syntactic), and is identified by "an immediate contrast between the child error and a correct alternative to the error, as supplied by the child's interlocutor" (Saxton, 1997).
As revealed by the description by Saxton (1997), the positive input is not sufficient enough for a child to be able to correct his mistakes; rather, negative data is needed as well.In fact, negative evidence is a kind of pre-requirement for the child in order to be able to get rid of mistakes.Regarding the scientific knowledge, they send to the child, negative sign, and active inputs are identical; nonetheless, they change only in the circumstances they are employed.Negative data is only directly utilized after the child makes a mistake.

Literature Review 2.
No agreement is reached among scholars on the effects of negative data in language finding.Negative evidence is entirely rejected by Nativists who say no one corrects child's mistakes (Flevaris, 2011).They believe that children's innate capacity as well as the frequency of grammatical statements can help them to cross out the ungrammatical sentences from grammatical sentences.Furthermore, they can guess possible grammatical structures.That is how child could quickly adjust their incorrect formations (Backley, & Gallaway, 2005).Morgan, Bonamo, and Travis (1995) argue that correct grammatical structures do not occur frequently enough in the input which are necessary for learning.Sometimes, modification may not be fruitful.
In other words, non-nativist proponents claim that negative data is pivotal in molding the child's language and changing it into a grammatical one.Clark and Chouinrad (2003) declare that contrary data is adequate to permit the child to fix his mistakes.Saxton et al. (2005) assume the presence of an internal device for language training; nevertheless, they do not fully reject the role of improving such as Nativists.Saxton (1997) presented data that the child's quick responses to negative data are often compatible with its restorative function.Saxton (1997) provided the following instances: 1) Child: That policeman falled all the way down to the tiger.Adult: He fell dow.Child: Yes, he did.He fell down cause he likes that tiger.2) Child: It's even gooder than anything.(repeated 4 times).
Adult: Yes, it's better.Child: Better, yeah.Saxton (1997) concludes that authentic data show children to alter from erroneous to improve grammatical formations following the interference of negative data.Saxton et al. (2005) assumed that negative data plays a significant role in language education and it is not in opposition to natural ability.They supposed that a productive information does not mean that there is any inherent capability for language training.As they put.
Inevitably, both could co-exist.Actually, nativist scientists profess the question "what would the child's endowment for language look like if negative input were not available?"In so doing, one is more likely to find the full characteristics and amount of natural structure in language learning.(Saxton et al., 2005).
Childs are exposed to various erroneous forms.When parents speak, they make slips of the tongue and grammatical errors.But, all Childs grew up knowing the language correctly.In Chomskyan paradigm, it is assumed that parents do not usually improve their children's grammatical mistakes they ordinarily edit the content of their children's expression preferably than their grammatical mistakes.Grammatical mistakes are only seldom corrected.Brown and Hanlon (1970) found that child's grammatical mistakes were not edited, and they could make a judgment separating grammatical and non-grammatical sentences according to their LAD.Their conclusions certainly present experimental support for Chomsky's argument from the lack of the incentive.If Childs do not get any guidance from their linguistic conditions, they should explicitly apply their internal potential.Saxton et al. (2005) asserted that improvement of children by parents consists of a form of negative data containing the data regarding the grammaticality and ungrammaticality of comments.Children are assaulted with a mild, fluent, grammatically right, and context-bound speech.Children find the right form of their intended purpose with the aid of their parents.Parents help their children in forming complete sentences, to paraphrase, or to repeat the right form of their statements (Pinker, 1984).Children can conclude the right form of language from their parents' speech; however, this child-directed speech differs from negative data (Snow, 1997).
Language acquisition can be controlled by negative input and this negative input helps Childs to get rid of their mistakes in the method of language acquisition.A valid case of this is where the child makes mistakes with past tense forms of irregular verbs.The child grammar at that time allows both 'writted' and 'written,' while adults' grammar only allows written.In some stages of language acquisition, as Saxton et al. (2005) declared, child's grammar allows 2 forms where the adult grammar permits only one.In this case, negative data assists the child to find out which one are essential (Saxton et al., 2005).Marcus (1993) stated that 'external clues' from people around and 'internal mechanism' are two methods to assist children in understanding his statements is not grammatical.Parents can explicitly deny child's ungrammatical statement, or the child should be implemented with an inherent tool to avoid ungrammatical statements.Based on the Saxton et al. (2005), negative data makes a novel connection for the child to recognize the discrepancy between his own statements and those presented by his parents.To get this contrast, the child requires two pre-requisites.Head, the child, should know that the form provided by the adult is grammatical.Next, he should realize that his own sentence is ungrammatical.Marcus (1993) introduced three kinds of negative data: full feedback, partial feedback, and noisy one.Full feedback is applied only for ungrammatical statements; it is not utilized for correct ones.Some ungrammatical statements support partial feedback; but, it is never employed after right statements.Noisy one is "a corrective signal provided after ISSN 2039-9340 (print) Mediterranean Journal of Social Sciences MCSER Publishing, Rome-Italy Vol 7 No 5 S1 October 2016 20 some errors and after some correct sentences, but in different proportions".
According to Chouinard and Clark (2003), there are various meanings of adult correction for kids: a) they made a mistake; b) the place of that mistake, and c) the right form to be applied alternately.In other words, by examining their own statement and that of adults, children will follow the place of the fault and "identify the conventional form that they should have used for that meaning" (Chouinard & Clark, 2003).Saxton (1997) said that "despite what other differences may exist between child and adult utterances, [children] adopt the form favored by the adult in place of their own version".Clark and Chouinard (2003) declared that Childs attend to the adults' improvements as they repeat the revised forms, and accept those revisions with responses like 'uh-huh' or 'Yeah.'Based on Bernicot and Clark (2008), seldom even if children do not get used of adults' traditional forms, they put them in their mind so that after temporarily they have the needed inherent to repair faults.Bates and Marchman (1994) assumed that child has some entrance stage of exposure since they completely adopt an adult form to substitute their own information.
As mentioned before, there are different points of view in the assumptions of nativist in assumptions of child language acquisition and the advocates of 'Contrast Method.'The former assume that adults do not fix the incorrect statements of their child and the children do not want any negative data to move to the high amount of support (Pinker, 1989).Nevertheless, the latter demands that children can differentiate among positive input and negative data.Additionally, they can respond to them individually.The effectiveness of negative data looks to be more than positive one because it stimulates the children to use parents' right forms (Saxton, 1997).
To the researchers' best experience, there seem to be few types of research on parents' sense in improving their children incorrect statements.Consequently, the researchers asked to review how Azeri parents in Urmia, Iran, frequently correct their kids.The focus of this study is on various sorts of parental editing of Iranian child's acquiring Turkish as their original language.Partners were deemed to answer the subsequent questions: 1) What is your strategy in correcting your child errors?
2) What kinds of errors are more common, grammatical, pronunciation or lexical? 3) Does each type of error lead to uptake?Method 3.

Participants
As it is common in Iran, children within age range of 6-7 attend pre-school or 'pish-dabestani' prior to attending the primary school.The pre-school can be considered as one of the most sensitive stages in education for almost all the parents and grandparents.Azeri families are generally considered as 'collectivist' families rather than 'individualist'.It follows that almost grandparents live in one house with their children and grandchildren, and when parents are out at work, grandparents are responsible for taking six-year old children to the schools.As a result, the samples of the present study were 70 parents as well as grandparents in Urmia as one of the major cities in Iran.The language of the people in Urmia is Azeri or Turkish in addition to standard language of the country, i.e.Farsi.It is worth mentioning that the language of majority is Azeri; however, minorities like Christians and Kurdish people who are living in the same context are different.For the aim of the present study, the researchers opted for just Azeri monolinguals rather than bilinguals that exposed their children to both languages of Azeri and Farsi.The parents and grandparents were between 22-58 years.Totally, 80% percent of the children were the first child and the rest of them were second or third children (about 20%).

Design
The design of the study was a descriptive approach.As Kahn and Best (2006) presented, the aim of the descriptive study is to understand answers to "analysis of past events or the already existing conditions".It defines and interprets whatever exists.An unstructured interview with parents and grandparents using three general questions about the method of error correction on the parents' turn, their reactions to the children' errors, and different types of errors with samples was the only instrument used in this research.Ex-post facto design was necessary for the aim of this study since the researchers had no handle over what has previously passed to the points of the existing research.

Data categorization and coding system
Regarding coding system, having listened to the recorded files, the researchers categorized errors into three parts based Mediterranean Journal of Social Sciences MCSER Publishing, Rome-Italy Vol 7 No 5 S1 October 2016 21 on their linguistic focus, i.e. grammatical, phonological, and lexical.This categorization is warmly and widely welcome among the scholars who are active in the field of error treatment.For example, Zhao and Bitchener (2007), in the area of error correction in the classroom, classified them like the current study; however, McCarty (2008) added L1 to the category and admitted it as an error since it followed teacher 'feedback'.The present study, on the other hand, had different view about L1 errors, and classified them under the subcategory of the lexical errors.The grammatical and the phonological errors were clear for the researcher, but lexical errors were vague and complex due to their nature.To eliminate this problem and to distinguish the lexical errors from other errors, the researcher decided to categorize them according to Yule's (1996) categorization of morphemes.Yule (1996) categorized morphemes into free and bound morphemes.Free morphemes contain lexical and functional morphemes.The former contains nouns, verbs, and adjectives which carry the content of the message.The latter contains conjunctions, pronouns, and articles that are called 'closed' class since they cannot be added new ones to the language; as a result, if the missing word was a functional word, it was categorized as grammatical error.However, if the missing word was a content word, it was coded as lexical error.The following example illustrates a lexical error: 3) Child: mama gosom var Parent: otorosson getsin Child: yokh mama gosom var Parent: didim otorosson.As it is clear, there is a problem between parent and child in understanding the meaning of 'vomit' and 'bird'.Comprehending the meaning in this case is problematic for them because the word is a lexical error.The child's meaning of 'gos' is 'vomit'; however, the parent thinks that she means 'gosh' or 'bird'.
Phonological errors include segmental and supra-segmental aspects of the phonological system (Zhao & Bitchener, 2007) that contain the pronunciation of the individual words.Following is an example of phonological error: 4) Child: mama mahigamiz harda?Parent: jan? lapigamiz?Child: ha mahigamiz?The bold words in the sentence indicate that in the first move the child mispronounce the word mahiga instead of lapiga.In the second move the parent directly correct the phonological word, but in the third move the child repeats the same error without uptake.

Data analysis
This study hypothesized that Turkish parents ignored their children' different types of errors like pronunciation, grammatical and lexical errors.In order to test this hypothesis, parents answered the first research question based on their memories from their children's' errors in the past years.The researcher asked them to bring examples from the errors and then subcategorized the parents' reaction to the errors into explicit correction, implicit correction, and ignorance (for the first research question).The second research question tried to investigate the types of errors, i.e. whether they were grammatical, pronunciation or lexical.The third research question related to the children's uptake.Uptake is a term that is more common in the classrooms; as a matter of fact, it is reaction to the feedback or representing correct form by the child after that parent signaled that there is something wrong with the sentence or phrase.The results are estimated numerically through the following formula: {X N}.100=y In which, X is the number of examples of analyzed category, N is the total number of the examples, and, y is the percentage of occurrence of that category.

4.
To answer the first query, i.e. what is your strategy in correcting your child errors?, parents brought examples of the correction and the researcher subcategorized them into explicit, implicit, and ignorance strategies.Each parent gave three examples and, as a result, 210 samples of errors, strategies as well as uptakes were gathered.From the total of the samples, 124 ignored the errors, 57 implicitly corrected the errors, and only 29 explicitly corrected the errors (see Table 1).As Table 1 shows, parents used the strategy of ignorance about 59%, while implicit and explicit strategies were used 27% and 14%, respectively.Therefore, the first hypothesis was accepted which claimed Azeri parents mostly ignored their children's errors.Figure 1 represents these results.As Table 2 represents, frequency and percentage of pronunciation errors were more than those of grammatical and lexical errors.Therefore, the second hypothesis was rejected which claimed lexical errors were the most common types of errors commited by Turkish children.

Figure 2. Percentage of different types of errors
Uptake was another important category that the researchers were interested in discovering through the examples that the parents were sure about them.The third hypothesis claimed that uptake rarely happens after the parents' error correction.In order to test this hypothesis, the extracts were carefully transcribed and analyzed.The results revealed that from the total of 210 errors, just 28 (13%) samples led to uptake and the rest, i.e. 182 (87%) samples of the errors, remained without corrections.As a result, it can be concluded that uptake was low in Azeri children's utterances.The third hypothesis was accepted which claimed uptake rarely happens after the parents' error correction.Table 3 and Figure 3 represent these results.Discussion and Conclusion 5.
The question regarding the existence and the impact of negative data looks to be an essential query in the area of L1 and second language (L2) acquisition, especially in L1 acquisition.The reason is that by denying negative data in reaction to child's mistake one will find the kid is relying on his inherent capability to train the language (Marcuse, 1993).There are very different points of view in the being useful of negative data.Saxton et al. (2005) claim that although negative data might facilitate language acquisition, it is an unnecessary component of language acquisition.Marcus (1993) argues although parents supply noisy feedback for improving children's mistakes, children rely on their internal mechanisms.In fact, it is not easy to believe children can use just these noisy feedbacks to correct their errors.He found that inner sense and device are essential conditions in order to account for uptake.However, behaviorisms and non-nativist scholars assume that child language learning happens with just input or the main source of all linguistic information in tabula rasa.They emphasize the role of negative data as a crucial role in reforming child's ungrammatical language into the grammatical one.
The outcomes of this research demonstrated that only 14% of Azeri parents corrected their children's mistakes explicitly.In addition, 27% of the parents corrected implicitly, and a big number of the participants ignored their children's errors because they were not reluctant to hear erroneous utterances from their own children who were welcome for the parents.The findngs also show parents are sensitive to the different types of mistakes children cause.They mostly correct phonological errors of their children.Moreover, uptake was low after parents' reaction to the children's ill-formed utterances.The findings of this study are to some extent in line with Birjandi and Nasrolahi's results (2012).In a descriptive study, Birjandi and Nasrolahi (2012) tried to find the manner in which Persian parents corrected their children's erroneous utterances.Similar to the present study, they found that a big amount of errors happened in the form of phonological errors; however, parents were sensitive to grammatical types of errors.Also, the present study was not in harmony with Birjandi and Nasrolahi's study.That is, while the adults in Birjandi and Nasrolahi's study expressed that they were utilized to Child's ill-formed utterances and even enjoyed mistakes their child make, regarding more than half of the parents (61%) corrected their children's errors.
Previous investigations have discovered that moms responded to their children's incorrect comments no more than 34 percent of the time (Hirsh-Pasek, et al., 1984;Farrar, 1992), which based on Flevaris (2011) is not sufficient for the elimination of mistakes.The verdicts of this research showed that Azeri parents do correct their children's ill-formed ISSN 2039-9340 (print) Mediterranean Journal of Social Sciences MCSER Publishing, Rome-Italy Vol 7 No 5 S1 October 2016 24 utterances and their correction seem to play the role of catalyzer for language learning; however, their reaction to the errors does not completely lead to high amount of uptake.The children use truth value and use their internal mechanism which can never be replaced by negative evidence, as Morgan et al. (1995) suggests.The role of mental mechanisms in the child's brain is ruling out the grammatical utterances produced by parents.As a result, this study's findings support Saxton's (1997) idea which claimed negative input in addition to positive data and inner device reshapes the kid language.According to Saxton (1997).
The finding that negative data represents a function in the improvement from mistakes would not permit one to assume that natural parts and positive data have no capacity in the removal from overgeneralization.In other words, it would not be unexpected to find that all three parts (positive and negative input and innate attributes) provide in some way.
It is critical to perceiving the constraints of this research.First, this investigation was conducted only in one city, i.e.Urmia, with only 70 participants.It is axiomatic that the greater sample size from different cites with differences in dialects can provide a better picture of error treatment of Iranian children.Also, variety in parents' socio-economic characteristics can significantly play a role in their perception of children's mistakes and their modification.The participants in the present research looks to be from the same socio-economic background.There is a need to conduct more studies in different socio-economical contexts, which could lead to a better picture of mistake correction perspective.

Figure 1 .
Figure 1.Percentage of strategies used by the parents The second research question attempted to answer the following question: 'what kinds of errors are more common, grammatical, pronunciation or lexical?'.From about 210 extracts of errors, about 57 (27%) extracts were grammatical, 23 extracts (11%) were lexical, and the rest of them, namely 130 extracts (62%), were pronunciation errors.Table 2 and Figure 2 represent the results.

Table 1 .
Frequency and percentage of strategies used by parents

Table 2 .
Frequency and percentage of different types of errors

Table 3 .
Frequency and percentage of uptakes