2.1 Learner's Language
Following the discovery of the weaknesses of the Contrastive Analysis Hypothesis (CAH) and the evolution of Error Analysis (henceforth EA), attempts were made at developing and understanding of the processes of second language learning. Emphasis was shifted from studying and analyzing the systems of the native and the target languages to the analysis of the learner's language which began to be seen as a phenomenon to be studied in its own right (Keshavarz, 1999: 55)
Researchers and teachers have come more and more to understand that second language learning is the process of the creative construction of a system in which learners are consciously testing hypotheses about the target language from a number of possible sources of knowledge: knowledge of the native language, limited knowledge of the target language itself, knowledge of the communicative functions of language, knowledge about language in general, and knowledge about life, human beings, and the universe (Brown, 2000: 215)
By the late 1960s and early 1970s, the second-language learner's errors gained unprecedented prominence and became the subject of rigorous investigation in their own right. Alongside this extended domain of EA, revolutionary concepts developed in the study of second language learner's language. The best known of these is ''interlanguage'', a language which is between two languages (the source and the target language). The term interlanguage was coined by Selinker in 1969 to refer to 'language-learner's language as a separate linguistic system based on the observable output which results from a learner's attempted production of TL norm' (Keshavarz,1999: 56).
William Nemser (1971: 116) employs the term ‘Approximative system’ to identify a learner’s linguistic system which is distinct from his/ her mother tongue and the target language he/ she is attempting to learn. Here Nemser's term ‘approximative’ means that the learner is progressing towards the target language and his system is developmental in nature. The term system implies that he is using a set of rules and hence his language is not a random.
As the learner receives more and more exposure and collects new data he attempts to change his system. He constantly tries to alter his system to bring it closer to the target language. Hence his system should be studied from three points of view (1) his mother tongue (2) his target language and (3) as an independent system itself (Nemser, 1971; reprinted in Richards, 1974: 56).
It is important to study this system separately because it can ‘provide attested information or immediate utility in teaching and course development on patterns of learning behavior for the principal structures of the target languages’ (ibid: 63). What Nemser is trying to suggest is that the materials based on contrastive linguistic studies are not so effective because they take into consideration only the learner’s mother tongue and his target language. If learner’s learning behavior as revealed from the study of his approximative system is understood, we should be able to foresee the problems of a particular learner with respect to a particular target language.
The main difference between the 'approximative system hypothesis' and the 'interlanguage hypothesis' is that the latter emphasizes the structurally intermediate status of the learner's language system between mother tongue and target language, while the former emphasizes the transitional and dynamic nature of the system (Keshavarz, 1999: 61-62).
Corder suggests that linguistics should study the process of second and foreign language acquisition and the various strategies learners may use. He has contributed a number of articles wherein he discusses the nature of the learners’ language. He calls the learners’ language as their idiosyncratic dialect (1971: 151). He says that this dialect of the learners is (1) regular (2) systematic and (3) meaningful.
According to him the learners’ utterances can be accounted for by a set of rules. This set of rules is obtained from the target social dialect. He gives two reasons why the learners’ language should be considered as a dialect of the target language. The two reasons are as follows:
(1) It is a language and has a grammar.
(2) At least some of the rules in this grammar are the same as those in the target language grammar (ibid).
Corder (1981, 66), points out that the study of the learners’ dialect would tell the teacher how far the learners have progressed towards the goal and what more they have to learn. He also points out that if learners utter a correct form we cannot take it as a proof that ‘the learners have learned the systems which would generate that form in a native speaker’, for they might just be repeating an utterance that they have heard before. They may not have understood the system behind it. In such cases they cannot be said to be using the language.
He also states that the child language and the language of aphasics are all deviant idiosyncratic dialects. Poetic language is ‘deliberately deviant’ and the language of the aphasics is ‘pathologically deviant’. But the dialects of the children and the learners are the result of the learning process. Here both the children acquiring their mother tongue and the learners learning a second language go through a similar process wherein they form hypotheses about the nature of the language and test them (Corder, 1981: 16-17). Corder, adds that an alternative name might be transitional dialect, emphasizing the unstable nature of such dialects (ibid: 18).
Corder (1967: 148), points out that an analysis of the learners’ language could help us adjust our syllabuses to the built-in syllabus which the learners have made for themselves. But it is not very easy to analyze the learners’ dialect, mainly because of two reasons. Firstly, the learners’ dialect is not stable and secondly, interpretation is difficult because of the peculiarity of the dialect. But if we understand the learners’ built-in syllabus through the study of their errors we could create better conditions for language learning.
To sum up, it can be stated that the basic idea which is common to all proposals is that the L2 learner constructs an internal grammar on the basis of the L2 input he receives: a grammar which is in subsequent stages or ''varieties'' keeps being reconstructed and will approximate a certain target variety of native speakers of that language more and more. This internal grammar is characterized most importantly by instability in the sense of change. The learner acquires and drops specific varieties continuously. In most L2 studies there are hardly any concrete hints as to how these changing learning varieties might be described.
A further complication is that language change over time does not exclusively manifest itself in the form of progress. In addition to progress one may find 'levelling', or even regression ('back-sliding') (van Els et al, 1984: 70). Selinker uses the term 'fossilization' for both levelling and regression (1972: 36).
In addition to the methodological problems already mentioned, differences in research design are found. These differences relate to choice of (1) period of time, (2) informants, and (3) data collection procedures. These are to be discussed in the following section.
2.2 Research Design
2.2.1 Type of research
First of all a choice has to be made between longitudinal and cross- sectional research designs. In the first type of research, the language behavior of one and the same informant or group of informants is registered for a certain period for specific intervals (e.g. one hour, twice a week, for 12 months). In the second case, one single sample of the language behavior of a group of informants is taken (van Els et al: 1984:70).
From such a sample one may draw conclusions about which aspects of L2 have been mastered, and to what extent. In contrast with cross-sectional studies, where the time factor has been eliminated, longitudinal studies give a picture of language development over time. The cross-sectional allows statements on the order of accuracy (or order of difficulty) of a series of linguistic units, whereas the longitudinal design allows statements on the order of acquisition. In many cross-sectional studies it is assumed that an observed order of accuracy reflects a non-observed order of acquisition (Fathman, 1977: 30). In other words, synchronic data are interpreted diachronically.
A lot of research into language development is cross-sectional, which has to do with the fact that longitudinal research is by definition time-consuming (and therefore often expensive), and with the difficulty in applying it to large numbers of informants. Ellis (1986: 58) points out that one of the disadvantages of the longitudinal research design lies in the difficulty of making generalizations based on the profiles of one or two learners. Thus, the present study is going to be of the cross-sectional approach.
Statements about language development can be based on observed behavior of one, some, or many informants (van Els et al, 1984: 71). There are several disadvantages in case studies which are not to be found in studies involving larger groups:
it is impossible to generalize about language development on the basis of case studies;
case studies do not allow for statements about individual variation in language development;
they can easily lead to observer bias: the researcher may identify with the informant or informants to such an extent that he observes what he wishes to observe;
frequent contact between researcher and informant which is aimed at collecting data on 'language' may influence the language development process in as yet unknown ways (van Els et al, 1984: 71).
Thus, and for the above mentioned reasons, the present study involves large groups of subjects (See Chapter Four).
2.2.3 Data Collection Procedures
Having adequate data is essential for any EA. The choice of appropriate procedures for collecting data is, in fact, one of the crucial steps in the investigation of the learner's language (Keshavarz, 1999: 65). There are many methods of data collection, ranging from observation of natural behavior (or, at least, behavior which is as natural as possible, since natural behavior is hardly ever observed) (van Els et al, 1984: 72).
Naturalistic observations are a rich source of expected and unexpected linguistic data. But, what to be noticed here is that by the word 'unexpected' it is indicated that such observations very often have the goal to generate hypotheses.
This actually highlights one of the arguments in favor of experimental tasks:
Experimental research usually tests rather than generates hypotheses; on the basis of well defined hypothesis-oriented tasks it is easier to collect, order, and interpret specific linguistic data.
One the one hand, experimental research may prevent crucial gaps in the data-base (there are no 'accidental' data, as with naturalistic observations), while on the other hand, it prevents the emergence of a vast range of 'useless data' (van Els et al, 1984:72).
Thus, and to get the use of the already mentioned points, the present study follows the experimental approach in its research.
These difficult choices of period of time, informants and procedures of data collection indicate that there is no such thing as a 'royal road' to insight into language development.
2.3 Error Analysis
2.3.1 Historical Background
Researchers have resorted to the outputs of learning processes of which the most observable manifestations are the errors committed by learners, because they have found themselves forced to cope with the problem of difficulty in L2 learning without adequate knowledge of learning mechanisms. To diagnose difficulties, learners' errors are systematically collected, analyzed, and tabulated. This approach is commonly known as EA, and is based on assumption that the frequency of errors is proportional to the degree of learning difficulty.
Brown (1987: 170) points out that researchers and teachers of second languages came to realize that the mistakes a person made in the process of constructing a new system of language needed to be analyzed carefully, for they possibly held in them some of the keys to the understanding of the process of second language acquisition. As Corder (1967: 167) has noted: ''A learner's errors …. are significant in [that] they provide to the researcher evidence of how language is learned or acquired, what strategies or procedures the learner is employing in the discovery of the language.''
Though EA was developed in the first place in the teaching of L1 between 1915 and 1933, studies of errors in L2 learning were scarce, and it was until the late 1950s that an increasing number of such studies were reported (Chau, 1975: 120).
The goals of EA during these early stages were only pedagogic: it consisted of lists of 'common' errors and their linguistic classification, which were used to provide the basis for the organization of teaching materials and the specification of remedial lessons and exercises (Ellis, 1986: 51).
The interest in EA, however, declined after the growing enthusiasm for Contrastive Analysis (henceforth CA) which was considered a more reliable method than EA for predicting and explaining learning difficulties (Chau, 1975: 123). The CA hypothesis, in its strong version has claimed that ''L2 learning problems can be predicted on the basis of the linguistic differences between L1 and L2'' (van Els et al., 1984: 50). Influenced by the Behaviorist learning theory, CA has claimed that the prevention of errors (the goal of CA) is more important than their identification (Ellis, 1986: 51)
However, at the end of 1960s EA regained its prestige after researchers had realized, on the basis of the empirical data provided by EA, the inadequacy of the CA hypothesis for predicting L2 learning problems. It has been found out that linguistic differences between L1 and L2 do not necessarily lead to L2 learning problems, and that not all L2 learning problems can be traced back to the linguistic differences between L1 and L2, for example some L2 learning problems are caused by the structure of L2 itself, and thus can not be predicted on the basis of CA (van Els et al., 1984: 49-52).
Furthermore, behavioral psychology, on which the CA hypothesis is based, has been strongly criticized on the grounds that it has left the language learner out of consideration, neglecting his creative contribution to language learning, and considering him merely a passive participant in the process of language learning, totally dependent on his teacher for drill and reinforcement in order to learn (Taylor, 1975: 392). These findings have resulted in more attention being paid to EA, and in maintaining a weakened version of the CA hypothesis, namely that '' some observed L2 learning problems can be explained on the basis of linguistic difference phenomena (Schachter, 1974: 206).
The most significant contribution of EA after the late 1960s, apart from the role it played in the reassessment of the CA hypothesis, lies in its success in elevating the status of errors from undesirability to that of a guide to the learner's cognitive strategies and his active contribution to the language learning process (Ellis, 1986: 53-54). Corder (1967: 168-69) stresses that errors are significant in that they are indispensable devices (strategies) a L2 learner (like a child acquiring his L1) uses to test his hypotheses about the nature of the language he is learning, and that they provide evidence of the system of the language he has learned or is using at a particular point in the course, i.e., his interlanguage.
Corder (1981: 1) points out that the field of error analysis can be divided into two branches: (i) theoretical, and (ii) applied. Theoretical analysis of errors, as mentioned before, primarily concerns the process and strategies of language learning and its similarities with first language acquisition.
In other words, it tries to investigate what is going on in the minds of language learners. Secondly, it tries to decode the strategies of learners such as overgeneralization and simplification, and thirdly, to go to a conclusion that regards the universals of language learning process whether there is an internal syllabus for learning a second language.
Applied error analysis, on the other hand, concerns organizing remedial courses, and devising appropriate materials and teaching strategies based on the findings of theoretical error analysis.
The procedures of EA as outlined by Corder (1974: 126-31), consist of a series of successive steps which involve in addition to the selection of a corpus of language to be elicited via elicitation techniques such as tests (See Chapter Four), recognition or identification, classification, and explanation of errors.
2.3.2 Errors in first and second language acquisition
Language learning is a process in which, like learning to swim, learners profit from mistakes by obtaining feedback to make new attempts that successively approximate the desired goals. Learners' errors can also ''provide to the researcher evidence of how language is learned or acquired, what strategies or procedures the learner is employing in the discovery of the language'' (Corder, 1967: 167). However, sometimes the feedback from an error can be so strongly negative that the learner would give up making new attempts and the learning would end up in failure.
During the trail-and-error process, the rules that a learner formulates are proved correct if the form he produces is acceptable in the L2, but need to be revised if the form is unacceptable. The latter appears as an error in his speech. If we apply this interpretation of the learning process to second language learning, we find that there is one substantial difference from the L1 learning situation. The child learning his L1 is exposed to one language only and can make his hypotheses about the rule structure on the basis of that language and whatever innate notions of language he may have (Wilkins, 1972: 203).
After 1960 more and more studies appeared in which the language behavior of children was characterized as the (subconscious) application of rule systems deviating from the rule systems which adult speakers use. Deviations from these norms started to be considered as inevitable, necessary, and systematic stages in the language learning process and are taken to constitute (subconscious) hypotheses by the child about the language to be learned (van Els et al, 1984: 49). In L2 learning, learners also regularly produce deviations from the L2 norm. It was generally felt that L2 learners would only learn what they were taught and would learn nothing that they were not taught. If errors occurred, in spite of the teaching, this was invariably attributed to interference from L1. At the end of 1960s, people began to question one of the main objectives of CA, namely the explanation and prediction of L2 learning problems. People began to realize more and more that this approach left the L2 learner out of consideration. The fact that there was no empirical basis for CA in turn resulted in more attention being paid to EA (van Els et al, 1984: 49).
In this way, errors of both L1 and L2 acquisition are seen to be the product of the same overall learning process. While providing a framework within which all the evidence of error can be discussed, it does not contribute to our understanding of why the learner bases his hypotheses on the mother-tongue at times and on the target language at other times. Nothing like enough research has been done into second language learning for us to be able to understand the interaction between these two ways of applying a learning strategy (Wilkins, 1972: 204).
2.3.3 Significance of Errors
Many scholars in the field of EA have stressed the significance of second- language learners' errors. Corder in his influential article (1967), remarks that …'' they are significant in three different ways. First, to the teacher, in that they tell him, if he undertakes a systematic analysis, how far towards the goal the learner has progressed and, consequently, what remains for him to learn. Second, they provide to the researcher evidence of how language is learnt or acquired, what strategies or procedures the learner is employing in his discovery of the language. Thirdly, they are indispensable to the learner himself, because we can regard the making of errors as a device the learner uses in order to learn. It is a way the learner has for testing his hypotheses about the nature of the language he is learning.''(Reprinted in Corder, 1981: 10-11).
Corder's views in this regard have been reiterated in the literature. Richards (1971a; cited in Keshavarz, 1999: 45), for example, remarks that errors are significant and of interest to:
Linguists, because as Chomsky suggests the study of human language is the most fruitful way of discovering what constitutes human intelligence.
Psycholinguists, because by looking at children's speech and comparing it with adult speech, they have been able to examine the nature of the mental processes that seem to be involved in language.
Teachers, because by analyzing learners' errors, they would be able to discover their difficulties and devise a method for comparing them.
Jain (1974: 208) also maintains that errors are significant for two reasons:
For understanding the process of L2 acquisition.
For planning courses incorporating the psychology of second-language learning.
2.3.4 Mistakes and Errors
Identifying an error goes beyond explaining what an error is. However, as linguists pay attention to the distinction between an error and a mistake, it is necessary to go over the definition of the two different phenomena.
According to Dictionary of Language Teaching and Applied Linguistics (2002: 295) a learner makes a mistake when writing or speaking because of lack of attention, fatigue, carelessness, or some other aspects of performance. Mistakes can be self-corrected when attention is called. Whereas, an error is the use of linguistic item in a way that a fluent or native speaker of the language regards it as showing faulty or incomplete
learning. In other words, it occurs because the learner does not know what is correct, and thus it cannot be self-corrected.
Mistakes must be carefully distinguished from errors of an L2 learner, idiosyncrasies in the language of the learner that are direct manifestations of a system within which a learner is operating at the time. An error, a noticeable deviation from the adult grammar of a native speaker, reflects the competence of a learner. Learners of English who ask, *''Does John can sing?'' are reflecting a competence level in which all verbs require a pre-posed do auxiliary for question formation. As such it is an error, most likely not a mistake (Brown, 2000: 217).
According to James (1998: 83) an error cannot be self corrected, while mistakes can be self-corrected if the deviation is pointed out to the speaker. But the learner's capacity for self-correction is objectively observable only if the learner actually self-corrects; therefore if no such self-correction occurs, so there will be no means to identify error vs. mistake.
Brown (1987: 171) considers the same point asking if we can turn to frequency of a deviant form as a criterion to differentiate an error from mistake. If on one or two occasions, an English learner says *''John cans sing'', but on other occasions says ''John can sing'', it is difficult to determine whether ''cans'' is a mistake or an error. If, however, further examination of the learner's speech consistently reveals such utterances as *''John wills go'', *''John mays come'', and so forth, we might safely conclude that ''cans'', ''wills'', ''mays'' and other such forms are errors indicating that the learner has not distinguished modals from other verbs. But it is possible, because of the few correct instances of production of this form, that the learner is on the verge of making the necessary differentiation between the two types of verbs.
The fact that learners do make errors, and that these errors can be observed, analyzed, and classified to reveal something of the system operating within the learner led to the increase of the study of learners' errors, called EA.
2.3.5 Preliminaries to the Analysis of Errors
There are three stages in EA: recognition, description, and explanation. These are logically dependent upon each other, and they will be discussed in the following sub-sections.
220.127.116.11 Recognition of Errors
One of the common difficulties in understanding the linguistic systems of language learners is the fact that such systems cannot be directly observed. They must be inferred by means of analyzing production and comprehension data. What makes the task even thornier is the instability of learners' systems, which are in a constant state of flux. When new information flows in it causes existing structures to be revised. Thus, in undertaking the task of performance analysis, the teacher and researcher are called upon to infer order and logic in this unstable and variable system (Keshavarz, 1999: 70).
Corder (1981: 21-22) remarks that the analysis of collected data involves several stages. The first stage in the technical process of describing the linguistic nature of errors is to detect and identify them. The difficulty in doing this, as he points out, lies in the fact that what looks or sounds a perfectly acceptable sentence may, nevertheless, contain errors. Of course, in many cases the sentence is not acceptable, i.e. the native speakers would not accept it as grammatical and then we know that it is erroneous. Such utterances, which are ungrammatical at the sentence level, are called overtly erroneous. Those which are erroneous, but not overtly so, are called covertly erroneous: sentences which are grammatically ' well-formed' at the sentence level but they are not interpretable in the context in which they occur. For example, ''Fine, thanks'' is grammatical and correct at the sentence level, but not when it is used in answer to ''Who are you?''.
The next step in the linguistic analysis of the collected data is to interpret what the learner has intended to say and to reconstruct his sentence in the target language. There are basically two types of interpretation: (a) authoritative and (b) plausible (Keshavarz, 1999: 71).
If we cannot interpret and describe a learner's errors and if we have access to the learner we can ask him to express his intention in his mother tongue, and then translate his utterance into the target language. This can be called an authoritative interpretation and an authoritative reconstruction of his utterance in an acceptable form (ibid).
However when the learner is not available we have to do the best we can to infer what he intended to say from his utterance, its context and whatever we know about his knowledge of the target language. If this plausible interpretation can be made of the sentence, then one should form a reconstruction of the sentence in the target language, compare the reconstruction with the original idiosyncratic sentence, and then describe the differences. If the native language of the learner is known, the model indicates using translation as a possible indicator of native language interference as the source of error. In some cases, no plausible interpretation is possible at all, and the researcher is left with no analysis of the error (Brown, 2000: 220).
18.104.22.168 Description of Errors
Corder points out that ''The description of errors is essentially a comparative process, the data being the original erroneous utterance and the reconstructed utterance''. Only the description which shows the respects in which the realization rules of the target language differ from those of the learner's dialect is of value. It is obvious that this can not be done unless we have adequate data. In other words, a single instance of an error is insufficient to establish that there exists a regularity (i.e. a set of rules) in the learner's dialect. It may represent merely a lapse or a mistake or a guess. It is only when we observe the same error occurring regularly that we can begin to talk about the rules the learner appears to be following and attempt a description of his transitional dialect (or that of the class as a whole). It is on the basis of systematic errors that we construct syllabuses and remedial programmes (1974: 128).
22.214.171.124 The Explanation of error
According to Corder (1974: 130-31), the making of errors is an inevitable, perhaps even a necessary, part of the learning process. It also accounts for the similarity of many errors to the forms of the mother tongue. These are called transfer errors. However, even when a learner has discovered a correct rule he may still continue to make errors because he has not yet discovered the precise set of categories to which the rule applies. Errors of this sort are errors of overgeneralization. It is clear that these do not necessarily have any connection with the nature of the mother tongue, and consequently we would expect to find that there is a set of errors made by learners of a particular L2 whatever their mother tongues. Thus, it is not surprising that errors like* be singed, *he cans come and *many mens are produced by learners with any mother tongue. There is a third type of error which is much more difficult to establish in any particular case, namely, errors arising from the methods or materials used in the teaching. It is, however, not easy to identify such errors except in conjunction with a close study of the materials and teaching techniques to which the learner has been exposed. This is probably why so little is known about them. Only this class of error, teaching- induced error, is avoidable or redundant and represents inefficiency in the learning-teaching process.
2.3.6 Sources of Errors
According to the CAH, the task of assigning errors to their sources was not a difficult one since the prime, if not the sole, cause of errors was interference coming from the learner's mother tongue. Such an analysis, of course, does not provide any account (apart from interference) about the psychological reality of errors, i.e. why learners commit errors, and what cognitive strategies and styles underlie certain errors. However one of the major contributions of the hypothesis of EA was its recognition of the sources of errors which extend beyond just interlingual errors in learning a second language. It is now clear that intralingual and developmental errors play an important role in second language learning (Cohen, 1975: 419).
Keshavarz states that in the late 1960s and early 1970s, empirical studies emerged with the realization that many kinds of errors, besides those due to mother tongue interference, were apparent in learner's languages (1999: 101). Observation such as those made by Duskova, (1969);
Buteau, (1970); and Richards, (1971b) initiated numerous investigations into sources of errors other than mother tongue interference.
Richards (1971a, reprinted in Richards, 1974: 173), for instance, points out that the limitation of certain strategies of rule learning gives rise to errors which are not caused by mother tongue interference, but by faulty application of learning strategies. The sources of these errors are discovered within the structure of the target language itself and some of them result from teaching techniques used. He calls these types of errors ''intralingual and developmental''.
Corder (1975: 203) distinguishes three types of errors with respect to their sources: (a) interlingual errors, which are caused by L1 interference; (b) intralingual errors, i.e. errors caused by the learner's generalizing and overgeneralizing particular grammatical rules; and (c) errors caused by faulty teaching techniques.
Dualy and Burt (1972: 238) categorize second language learners' errors, which they prefer to call 'goofs', as: Interference – like goofs, developmental goofs, ambiguous goofs, and unique goofs.
It should be stated that in the majority of cases an error may be attributed to more than one cause. Thus, the classification here is based on the primary causes of errors. That is, an error may primarily be attributed to one source while other causes may also be involved.
126.96.36.199 Interligual Transfer
Interlingual transfer is a significant source for language learners' errors. Dictionary of Language Teaching and Applied Linguistics (2002: 239) defines interlingual errors as being the result of language transfer, which is caused by the learner’s first language. Error analysis does not regard them as the persistence of old habits, but rather as signs that the learner is internalizing and investigating the system of the new language.
Keshavarz (1999:102) states that interlingual errors may occur at different levels such as transfer of phonological, morphological, grammatical and lexica-semantic elements of the native language into the target language.
The beginning stages of learning a L2 are characterized by a good deal of interlungual transfer from the L1. In these early stages, before the system of the L2 is familiar, the L1 is the only linguistic system in previous experience upon which the learner can draw (Brown, 2000: 177). For example, learners may say sheep for ship, or the book of Jack instead of Jack's book. According to Brown, these errors are attributable to negative interlingual transfer.
188.8.131.52 Intralingual Transfer
Interferences from the learners’ own language is not the only reason for committing errors. Brown points out that researchers have found that the early stages of language learning are characterized by a predominance of interference (interlingual transfer), but once learners have begun to aquire parts of new system, more and more intralingual transfer – generalization within the target language – is manifested (2000: 224). As learners progress in the second language, their previous experience and their existing subsumers begin to include structures within the target language itself (ibid).
The early stages of language learning are characterized by a predominance of interlingual transfer, i.e., interference, but once learners begin to acquire parts of the new systems, more and more intralungual transfer―overgeneralization within the target language―is manifested (Brown, 1987: 178).
This is justified by Brown as following logically from the tenets of learning theory. As learners progress in the L2, their previous experience and their existing subsumers begin to include structures within the target language itself (ibid).
Overgeneralization errors in the utterances of the language learner suggest active participation on his part (a risk-taking strategy according to Corder, 1978: 88) in the learning process by exercising his interlanguage creatively instead of imitating what he hears around him or transferring a L1 structure.
184.108.40.206 Context of Learning
A third major source of error, though it overlaps both types of transfer, is Brown's context of learning (Transfer of Training, Selinker, 1972: 35; Sociolinguistic Situation, Richards and Sampson, 1974: 6ff). Brown (1987: 179) explains that context refers, for example, to the classroom with its teacher and its materials in the case of school learning or the social situation in the case of untutored L2 learning. In a classroom context the teacher or the textbook can lead the learner to make faulty hypotheses about the language, what Richards (1971a: 178) called ''false concepts'' and what Stenson (1974: 58) termed ''induced errors'', including a) misleading explanation from the teacher, b) faulty presentation of a structure in a textbook, c) improperly contextualized pattern, d) confused vocabulary items because of contiguous presentation, and e) inappropriately formal forms of language----bookish language.
220.127.116.11 Communication Strategies
Communication strategies are a fourth source of learners' errors, they actually include processes of interlingual and intralingual transfer and the context of learning as a learner tries to get a message across to a hearer or reader. The communicative strategies pertain to the conscious employment of verbal or nonverbal mechanisms for communicating an idea when precise linguistic forms are for some reason not available to the learner at that point in communication (Brown, 1987: 180). Fearch and Kasper (1983: 36) define communication strategy as ''potentially conscious plans for solving what to an individual presents itself as a problem in reaching a particular communicative goal.''
Lengo states that in order to overcome a communication problem caused by a lack of or in ability to access L2 knowledge, learners have been found to resort to certain communicative strategies as a means of coping with the new situations. Thus, errors due to communicative strategies reflect gaps in linguistic knowledge or defective learning (1995: 23). Brown accounts for these strategies, among which avoidance is considered to be the most common one (1987:180).
Avoidance, according to Brown (1987: 183), is a common strategy that can be broken down into several subcategories and thus distinguished from other types of strategies. The most common type of avoidance strategy is syntactic or lexical avoidance within a semantic category. For example in the following conversation:
L: I lost my road.
NS: You lost your road?
L: Uh……. I lost. I lost. I got lost. (ibid: 184).
The learner avoided the lexical item road entirely not being able to come up with the word way at that point.
A more direct type of avoidance is topic avoidance, in which a whole topic of conversation might be avoided entirely. Learners manage to devise ingenious methods of topic avoidance: changing the subject, pretending not to understand, simply not responding at all, or noticeably abandoning a message when a thought becomes too difficult to continue expressing (ibid).
Avoidance, according to Corder (1978: 90), falls under message adjustment strategies, or risk-avoidance strategies. He states that among message adjustment strategies we have at one extreme topic avoidance, ''a refusal to enter onto or continue a discourse within some field or topic because of a feeling of total linguistic inadequacy''. A less extreme form of topic avoidance, as Corder believes, would be message abandonment: ''trying but giving up''. A less acute form of message adjustment is semantic avoidance that is 'saying something slightly different from what you intended but still broadly relevant to the topic of discourse'. Finally, the least acute form of message adjustment would be message reduction that is ''saying less, or less precisely what you intended to say''.