Jan. 2008, Volume 6, No.1 (Serial No.

52)

US-China Foreign Language, ISSN1539-8080, USA

Error theories and second language acquisition

XU Jie1
(School of Humanities, Hefei University of Technology, Hefei 230009, China)

Abstract: This paper critically examines in detail the three most influential error theoriescontrastive analysis, error analysis, and interlanguage theoryand evaluates their impact on SLA respectively. In light of these theories, this paper also probes into their enlightenment on foreign language teaching, and provides several suggestions about strategies for conducting second language teaching. The paper argues that an integrated approach, which combines three theories, is needed to deal with the complexities of SLA and provide empirical evidence for the improvement of teaching methodology. Key words: error; contrastive analysis; error analysis; interlanguage; second language acquisition

1. Introduction
Errors are an integral part of language acquisition. The phenomenon of error has long interested SLA researchers. In a traditional second language teaching situation, they are regarded as the linguistic phenomena deviant from the language rules and standard usages, reflecting learners deficiency in language competence and acquisition device. Many teachers simply correct individual errors as they occur, with little attempt to see patterns of errors or to seek causes in anything other than learner ignorance. Presently, however, with the development of linguistics, applied linguistics, psychology and other relevant subjects, peoples attitude toward errors changed greatly. Instead of being problems to be overcome or evils to be eradicated, errors are believed to be evidence of the learners stages in their target language (TL) development. It is through analyzing learner errors that errors are elevated from the status of undesirability to that of a guide to the inner working of the language learning process (Ellis, 1985, p. 53). In the field of SLA, there have been three influential approaches to errors with a general movement from approaches emphasizing the product, i.e. the error itself, to approaches focusing on the underlying process under which the errors are made. This paper is intended in the following sections to overview the basic assumptions of these three error theoriesContrastive Analysis, Error Analysis and Interlanguage Theoryand evaluates their application in SLA.

2. Error theories and their impact on SLA

2.1 Contrastive analysis and SLA 2.1.1 Contrastive analysis
This article is part of the authors ongoing research project A study of EFL learners errors which is sponsored by the Scientific Research Fund of Hefei University of Technology (No.071710F). XU Jie (1980- ), female, M.A., lecturer of School of Humanities, Hefei University of Technology; research field: applied linguistics. 35

Error theories and second language acquisition

In the 1950s and 1960s the favored paradigm for studying FL/SL learning and organizing its teaching was Contrastive Analysis (James, 2001, p. 4). Contrastive linguistics has been defined as a subdiscipline of linguistics concerned with the comparison of two or more languages or subsystems of languages in order to determine both the differences and similarities between them (Fisiak, 1981, p. 1). Johansson and Hofland held that language comparison is of great interest and can be categorized in a theoretical as well as an applied perspective (Johansson& Hofland, 1994, p. 25). Therefore, the study may be theoretical, without any immediate application, or may be applied, namely, carried out for a specific purpose. The applied contrastive studies, or alternatively Contrastive Analysis (CA) as was termed by Benjamin Lee Whorf in 1941, or Contrastive Analysis Hypothesis (CAH) by Wardhaugh (1970), belong within the scope of applied linguistics. CA stresses the influence of the mother tongue (MT) in learning a second language in phonological, morphological, lexical and syntactic levels. It holds that L2 would be affected by L1. Here, language is taken as a set of habits and learning as the establishment of new habits, a view sprung from behaviorism, under which language is essentially a system of habits. In the course of language learning, L1 learning habits will be transferred into L2 learning habits. Therefore, in the case of L1 transfer into L2, if structures in the MT have their corresponding structures in the TL and L1 habits can be successfully used in the L2, learners would transfer similar properties successfully and that would result in positive transfer. Contrastingly, in the case of negative transfer or interference, certain elements of the MT have no corresponding counterparts in the TL, L1 habits would cause errors in the L2, and learners would transfer inappropriate properties of L1. Two versions of CA were proposed, a strong version and a weak version, and on the former those who write contrastive analyses usually claim to base their work. Purists of contrastive analysis advocate a strong approachpredictions about learner difficulties and development of teaching methods based on a comparison of phonological, grammatical, and syntactic features of the NL and TL. A second or weaker version looks for learners recurring errors and attempts to account for those errors by ascribing their NL/TL differences. 2.1.2 Evaluation of contrastive analysis in SLA From the 1940s to the 1960s, contrastive analyses were conducted to investigate learner errors in the field of SLA, in which two languages were systematically compared. Researchers at that time were motivated by the prospect of being able to identify points of similarity and difference between NL and TL. There was a strong belief that a more effective pedagogy would result when these were taken into consideration. However, although CA was very influential and had inspired many productive investigations, its positions about its predictive power and the relationships between L1 and L2 learning faced serious challenges in the field of SLA. The original weakness of CA was its failure to go beyond a statement of difference to a supportable theory of difficulty. As Briere (1968) showed, difference by itself does not predict difficulty; often there is more difficulty in practice with similar structures than with different structures. Secondly, by viewing errors simply as a result of L1 interference, CA places the environment as the predominant factor in SLA, while learners are believed to play only a passive role in accepting the impositions of the environment. Thirdly, interference or transfer from L1 is not the sole source of errors in L2 learning, for a good number of the errors made by language learners seem to be unrelated to the learners native language. SLA researchers found such non-interference errors more pervasive in learner performance than contrastive analysts were ready to recognize. For example, Dulay and Burt (1973) studied the errors made by Spanish-speaking children learning English as an L2 and claimed that of all the learner
36

Error theories and second language acquisition

errors they had collected, 85% were developmental (i.e., non-interference errors), 12% were unique and only 3% were results of L1 interference. Under this circumstance, attempts were made to restore a refined contrastive analysis to the center of attention. Wardhaugh (1970) proposed a distinction between the strong and the weak version of CA. The latter method is believed to be the more valid of the two, and is certainly a more realistic and practical approach to detecting the source of error. Eventually, this weak, explanatory methodology became known as error analysis. 2.2 Error analysis and SLA 2.2.1 Error analysis EA received considerable attention and finally became a recognized part of applied linguistics in the 1970s since the strong version of CA turned out not to be a productive pedagogical tool. James defined the notion of EA as the study of linguistic ignorance, the investigation of what people do not know and how they attempt to cope with their ignorance (James, 2001, p. 62). It was Corder who made the first argument for the significance of learners errors in his 1967 seminal paper. The significance of learners errors, which signaled the shift of pedagogical interest from contrastive analysis to error analysis and provided the impetus for many empirical studies. In order to analyze learners errors in a proper perspective, EA enthusiasts considered it crucial to make a distinction between mistake and error, which are technically two very different phenomena (Brown, 1994, p. 205). Corder (1967) made use of Chomskys the competence versus performance distinction by associating errors with failures in competence and mistakes with failures in performance. In his view, a mistake occurs as the result of processing limitations rather than lack of competence. That is, it signifies L2 learners failure of utilizing their knowledge of a TL rule. All people make mistakes, in both native and second language situations. As a matter of fact, falling back on some alternative, non-standard language uses like false starts, hesitations, random guesses, confusions of structure or slips of the tongue is a regular feature of native speaker speech. Native speakers are normally capable of recognizing and correcting such mistakes. Nevertheless, an error, in this technical sense, is the breaches of rules of code; it is the noticeable deviation in grammaticality resulting from a lack of requisite knowledge. It arises because of the lack of competence. Native speakers may also make errors but they are able to correct their own errors; nevertheless, L2 learners cannot, by any means, always do so. The analysis of error sources has been regarded as a central aspect in the study of learner errors. Researchers believe that the clearer the understanding of the sources of learners errors, the better second language teachers will be able to detect the process of L2 learning. As already discussed above, it is competence errors that have been considered to be central to the study of SLA. In his A noncontrastive approach to error analysis, Richards (1971) identified a number of different sources or causes of competence errors: interference errors of MT interference, intralingual errors within the TL itself and developmental errors, reflecting the learners attempts to construct hypotheses about their target language from their limited experience. Excluding interference errors from his discussion, Richards (1971) focused on the intralingual and developmental errors observed in the acquisition of English as a second language and further classified them into four categories: (1) Overgeneralization, covering instances where the learners create a deviant structure on the basis of his experience of other structure of the TL; (2) Ignorance of rule restriction, occurring as a result of failure to observe the restrictions or existing structures; (3) Incomplete application of rules, arising when the learners fail
37

Error theories and second language acquisition

to fully develop a certain structure required to produce acceptable sentences; (4) False concepts hypothesized, deriving from faulty comprehension of distinctions in the TL. EA is a systematic study and analysis of errors made by the learners of a foreign language in an attempt to account for their origin, their regularity, their predictability and variability. It views both first and second language acquisition as a process involving the active participation of the learners. In this approach, errors are seen as a natural phenomenon that must occur when learning the first or second language before correct language rules are completely internalized. Errors occur systematically in learners language behavior and are, therefore, to be regarded as manifestations of an inner-working system 2.2.2 Evaluation of error analysis in SLA As was pointed out by some researchers (Corder, 1967; DAI Wei-dong, SHU Ding-fang, 1994; CAI Long-quan, 2000), by investigating leaner errors, EA can be highly significant to SLA in the following aspects: (1) They tell the teachers how far towards the goal the learners have progressed and what remains for them to learn. Errors provide feedback. They tell the teachers something about the effectiveness of their teaching materials and teaching techniques and show them what parts of the syllabus they have been following have been inadequately learned or taught and need further attention. In this way, they can provide learners with some more individual help and more appropriate tools depending on their specific needs and difficulties. Indeed, the attempt to discover more about L2 acquisition through the study of errors is itself motivated by a desire to improve language pedagogy. (2) They provide to the researchers evidence of how language is learned or acquired. Corder (1967) proposed as a working hypothesis that some of the strategies adopted by the learners of a second language are substantially the same as those by which a first language is acquired. By classifying the errors that learners made, researchers can learn a great deal about the SLA process by inferring the strategies that L2 learners are adopting. (3) They are means whereby learners test alternative hypotheses about the L2. For learners themselves, errors are indispensable, since the making of errors can be regarded as a device the learners use in order to learn. They is a way of testing hypotheses about the nature of language they are learning (Corder, 1967; DAI Wei-dong, SHU Ding-fang, 1994; CAI Long-quan, 2000; Selinker, 1992). Although EA has produced a far-reaching impact on SLA in many aspects, interestingly enough, EA, proposed as a remediation for the CA, also has defects. Till the seventies, it was very much criticized by researchers, who considered, among other things, that EA was besieged with methodological problems. Since it used to be difficult to collect or access large databases of learners language, a robust error typology that covers almost all error types was not established in traditional EA. Furthermore, EA does not shed much light on the developmental route learners take, because it examines languagelearner language at a single point in time. The real significance of EA cannot be ascertained without using diachronic data to describe learners developmental stages. The types and frequencies of leaner errors change in each acquisition phase. Without longitudinal data of learner language, it is difficult to yield a reliable result by EA. Another charge which has been leveled against EA is that it provides only a partial picture of what happens
38

Error theories and second language acquisition

when one learns L2. It can only deal effectively with the learner production (speaking and writing) but not with his perception (listening and reading), and it focuses on where the learner errs but ignores where he or she performs correctly. What is more, EA fails to account for the strategy of avoidance. For Schachter (1974), the fundamental flaw in EA was that learners do not often commit the expected errors because they tend to avoid words or structures they are not sure about. 2.3 Interlanguage theory and SLA 2.3.1 Interlanguage theory The concept of interlanguage (IL) was suggested by Selinker (1972) in order to draw attention to the possibility that the learners language can be regarded as a distinct language variety or system with its own particular characteristics and rules. IL is a structured and interlocking system which the learner constructs at a given stage in his development. An L2 leaner, at any particular moment in his learning sequence, is using a language system which is independent of both the TL and the learners MT. It is a third language, with its own grammar, its own lexicon and so on. The rules used by the learner are to be found in neither his own MT, nor in the TL. The earliest formulation of the notion interlanguage was that proposed by Corder (1967). One of its crucial contributions was its underlying assumption that the learners knowledge is to be seen as a unified whole, in which new knowledge is integrated and systematically reorganized with previous knowledge of the native language. By a gradual process of trial-and-error or hypothesis testing, learners slowly and tediously succeed in establishing closer approximations to the system used by the native speaker of the language. Various alternative terms have been used by different researchers to refer to the same phenomenon as IL. Corder (1971) proposed the notions of idiosyncratic dialects (also see Ellis, 1985, p. 47; Brown, 1994, pp. 203-204; Freeman & Long, 1994, pp. 60-61) to identify the idea that the learners language is peculiar and transitional competence to pinpoint the dynamic nature of the learners developing system. In another similar model, a paper by Nemser (1971) referred to this learner language as approximative system, one of a series of approximative stages through which the leaner moves in his acquisition of the TL. IL may be viewed as an adaptive strategy by which the learners try to construct the structural properties of the TL. This strategy uses simplification, reduction, overgeneralization, transfer, formulaic language, omissions, substitutions and restructurings (Selinker, 1972). Learners do not progress from zero knowledge of a TL rule to perfect knowledge of the rule. By using the above mentioned devices, they progress through a series of interim of developmental stages on their way to TL proficiency. These interconnected stages or systems form what Corder (1967) called the learners built-in syllabus (i.e. the interlanguage continuum). The form which IL takes can be accounted for by reference to a number of cognitive processes, five of which Selinker (1972) isolates as being of central importance in the language acquisition puzzle: language transfer, transfer of training, strategies of second language learning, strategies of second language communication and overgeneralization of TL linguistic material. The five processes together constitute the ways in which the leaner tries to internalize the L2 system. The characteristics of IL are described by many researchers as follows: 1) permeable, in the sense that rules that constitute the learners knowledge at any one stage are not fixed, but are open to amendment(Ellis, 1985, p. 50); 2) dynamic, in the sense that L2 learner slowly revises their variable interim systems to accommodate new hypothesis about the TL system; 3) systematic, in that L2 learners IL is
39

Error theories and second language acquisition

rule-governed, that is, the learner bases his performance plans on his existing rule system much the same way as the native speaker bases his plans on his internalized knowledge of the L1 system. 2.3.2 Evaluation of interlanguage theory in SLA Initiated in the early 1970s, IL studies view IL as a constant mental reorganization of a learners own version of the language and attempt to understand the systematic features of SLA. Since then, interlanguage has come to characterize a major approach to second language research and theory. Close analysis of the interlanguage development of individual learners has led to a better understanding of the nature and process of SLA. The Interlanguage theory realizes a complete shift of the focus from teaching perspective to learning perspective in SLA. Embracing less interest in the exploration of language pedagogy as traditional SLA approach did, interlanguage theory concerned primarily with the performance of L2 learners and established this performance as a suitable object of research. L2 Learners were considered to be playing an active role in L2 learning process. Explanations were sought in the learners mental processes that the learners used to convert input into knowledge and the knowledge systems which they construct mentally and manifest in output. The IL paradigm is also beset with problems. Several important criticisms about the IL studies should here be mentioned: (1) The concentration on morpho-syntactic development and the failure to deal with semantic development (Ellis, 1982). Interlanguage study is mostly limited to the scope of morpheme and syntax. The important aspect of meaning is rarely tackled. (2) The failure to define the concept clearly. Spolsky raised a problem with the notion of interlanguage, which was the tendency to confuse a process with a competence model (Spolsky, 1989, p. 33). Selinker seems to prefer a processing model in spite of his use of competence terminology. (3) The failure to develop effective approaches to facilitate empirical studies. The research methods of interlanguage study such as longitudinal and cross-sectional studies are technically inadequate in themselves. New and complimentary methods are yet to be found to probe into the nature and underlying principles of the L2 learners interlanguage.

3. Pedagogical implications
3.1 Contrastive analysis approach One of the strategies typical of L2 learners is to refer to their native language judiciously, and then make effective cross-linguistic comparison at different stages of language learning. Therefore, some information about contrastive studies of the two languages is needed so as to help students see more clearly some of the problems they may encounter. If learners are fully aware of all the potential problematic areas, they may successfully avoid making transfer errors, i.e., the third person singular simple present tense marker -(e)s to Chinese speakers of English. Effective learning of a second language should be based on the scientific perception of TL rules on the one hand, and on appropriate generalization of the native language rules on the other. Learners should build up their TL knowledge on the basis of the differences between L1 and L2 through comparisons of enormous details of the two languages.

40

Error theories and second language acquisition

3.2 Increase of target language input and output Reading and listening in the SLA situation is understood as the appropriate input for acquisition of language skills. Extensive reading and listening of authentic materials and wide exposure to meaningful and authentic TL contexts might provide input of linguistic knowledge at various levels, lexical, syntactic, and even rhetorical and cultural, to L2 learners. While being exposed to huge amounts of authentic TL input, learners should also be provided with ample opportunities to use and promote TL output. Producing the target language can provide students the opportunity to test out hypotheses about the TL and to move from a purely semantic analysis of the language to a more general understanding of it. 3.3 Effective feedback to learner errors Teachers are supposed to provide adaptive, qualified and motivating support for the individual learner in his learning process. It is important for teachers to be aware of the basic position that errors, as a natural and indispensable part of the learning process, should neither be tolerated nor corrected excessively. On the one hand, too much tolerance of learners errors may increase the risk of fossilization in the learners interlanguage. An IL can fossilize at a particular stage without ever becoming completely L2-like. This phenomenon of the non-progression of learning despite continuous exposure to input and opportunity to practice has been a central concern for SLA research. If the fossilized items, rules and subsystems are kept in the learners interlanguage, they will regularly emerge in the learners productive performance. This behavioral reappearance will of course impede the learners learning progress. On the other hand, it has been proved unreasonable to take all errors as signs of learners failure to learn. Rigid and immediate correction of every error at the time production may interfere with and alter learners path of thinking and induce severe psychological frustration. Therefore, in some cases, teachers may note errors and deal with them later, either at the end of the task, lesson, day, or in a following lesson. This non-immediate correction can also provide time for the teachers to research efficient and effective practice tasks to drill the learners in the correct forms.

4. Concluding remarks
In summary, although, from what has been discussed above, we can find Contrastive Analysis Hypothesis and Error Analysis have been overshadowed by Interlanguage theory in SLA, their value of studying L2 learning process and their significance in directing SLA research cannot be underestimated. Each of these three theories has devoted significantly to the study of SLA process. They can be looked upon as three evolutionary goals of one goal: to understand and explain the nature of the TL learners performance. Therefore, an integration of three schools is needed to deal with the complexities of SLA and provide empirical evidence for the improvement of teaching methodology.
References: Briere, E. 1968. A psycholinguistic study of phonological interference. The Hague: Mouton. Brown, H. 1994. Principles of language learning and teaching. Englewood Cliffs: Prentice Hall Regents. CAI Long-quan. 2000. The study of learning. Shanghai: Shanghai Foreign Language Education Press. (in Chinese) Corder, S. P. 1967. The significance of learners errors. International Review of Applied Linguistics, 5, 161-170.

41

Error theories and second language acquisition Corder, S. P. 1971. Idiosyncratic dialects and error analysis. International Journal of Applied Linguistics, 9, 149-159. DAI Wei-dong, SHU Ding-fang. 1994. Some research issues in contrastive analysis, error analysis and interlanguage. Journal of Foreign Languages, 5, 1-7. Dulay, H. & Burt, M. 1973. Should we teach children syntax. Language Learning, 23, 245-258. Ellis, R. 1982. The origins of interlanguage. Applied Linguistics, 3, 207-223. Ellis, R. 1985.Understanding second language acquisition. Oxford: Oxford University Press. Fisiak, J. 1981. Contrastive linguistics and the language teacher. Oxford: Pergamon Press. Freeman, D. & Long, M. 1994. An introduction to second language acquisition research. Essex: Longman. James, C. 2001. Errors in language learning and use: Exploring error analysis. Beijing: Foreign Language Teaching and Research Press. Johansson, S. & Hofland, K. 1994. Towards an English-Norwegian parallel corpus. In: Fries, U., Tottie, G. & Schneider, P. (Eds.), Creating and using English language corpora. Amsterdam and Atlanta, GA: Rodopi. Nemser, W. 1971. Approximative systems of foreign language learners. International Journal of Applied Linguistics, 9, 115-123. Richards, J. C. 1971. A Non-contrastive approach to error analysis. English Language Teaching Journal, 25, 204-219. Schachter, J. 1974. An error in error analysis. Language Learning, 24, 205-214. Selinker, L. 1972. Interlanguage. International Journal of Applied Linguistics, 10, 209-231. Selinker, L. 1992. Rediscovering interlanguage. London: Longman Group U.K. Limited, Essex. Spolsky, B. 1989. Conditions for second language learning. Oxford: Oxford University Press. Wardhaugh R. 1970. The contrastive analysis hypothesis. TESOL Quarterly, 4, 123-130.

(Edited by Stella and Katrina)

42