This paper tends to have a review of concepts and theories concerning error analysis and presents types of errors. It focuses on error analysis and studies written errors through subcategorization frame in Iranian EFL learners in elementary and intermediate levels. Seventy students from different instituts are asked to translate some sentences into English. The aim of this study is to find out whether lack of L2 subcategorizational information will cause errors in production of L2 and if increasing proficiency level eliminates the errors. Findings of this study is supposed to confirmed that lack of L2 subcategorizational information would be a major cause of errors which are highly affected by proficiency level.
One of the joys of EFL teachers is to see their students produce flawless English. Unfortunately it is inevitable to avoid errors of different types. Familiarity with the types of errors students actually make is a valuable guide for determining the sequence and emphasis of instruction in the EFL classroom. In recent years the acceptance of this notion has led to a great deal of empirical research on adult foreign language errors (Burt, 1975).
Error analysis was conceptualized and applied based on the behaviorist theory of language learning which implied that errors were signs that a language learner had simply not learnt the rules of the target language effectively (Brown, as cited in Khodabandeh, 2007). In the early 1950's, the notion of language as a system, and more importantly, the notion of second language acquisition as the meeting of two language systems gained more acceptance and linguists began to regard errors as evidence of language transfer, or what referred to as intersystemic interference. With this conceptualization, errors were regarded as the manifestation of ineffective language learning and were focused upon by linguists and teachers intent on their elimination (Khodabandeh, 2007).
It wasn't until the late 1960's that researchers like Slamecka and Ceraso (1960) used evidence gained through error analysis to discredit the existence of negative transfer as the dominant factor in acquiring a second language, since many errors could not be attributed to intersystemic interference. What they and other researchers were demonstrating was that learners didn't just memorize target language rules and use them to form their own utterances: they were constructing their own rules based on the input they had received. Hence, there was a rebirth of error analysis and a movement from an undifferentiated world to a world organized by mind, from a world of instances to a world related by generalities and abstractions requiring the whole concept to be redefined and approached from a more cognitive or mentalist perspective (Khodabandeh, 2007).
Psychology became a much more influential field for linguists interested in using error analysis as a diagnostic tool, to help identify the causes of errors. The term error itself was redefined in recognition that many mistakes in spontaneous speaking or writing could be attributed to a simple pause, metanalysis, or a "slip of the brain." (Crystal,as cited in Khodabandeh, 2007). Errors began to help describe and explain the way in which learners learned a language rather than their progress towards conforming to a set of real or imagined standards of expression and thus, had a more positive role.
Today, error analysis is used with a variety of techniques for identifying, classifying and systematically interpreting the mistakes made by language learners and has helped support hypotheses such as the natural route of development, as well as identify the weaknesses and/or disprove theories of language learning like contrastive analysis, throughout the last few decades (Khodabandeh, 2007). "The weaknesses of contrastive analysis are that it overemphasized the interference of the outer environment of language study, but the language learners themselves are totally neglected. While interlanguage intended to explore learning strategies based on the learners' errors, and it has become the basis of error analysis" (Fang & Xue-mei, 2007, p. 11).
Error analysis emphasizes "the significance of errors in learners' interlanguage system" (Brown, as cited in Khodabandeh, 2007). The term interlanguage, introduced by Selinker (1972), refers to the systematic knowledge of an L2 which is independent of both the learner's L1 and the target language. Nemser (1971) referred to it as the Approximate System, and Corder (1967) as the Idiosyncratic Dialect or Transitional Competence (as cited in Khodabandeh, 2007). "Error analysis is a type of linguistic analysis that focuses on the errors learners make. It consists of a comparison between the errors made in the Target Language (TL) and that TL itself" (Khodabandeh, 2007, p. 8). Burt, (1975) stated that "this 'error analysis approach' differs from that of 'contrastive analysis' in that error analysis does not assume that first language interference is the major predictor of adult errors. Rather, no assumptions are made about the causes of error types" (p. 54). Errors in the actual speech of foreign language learners were simply collected, then classified into categories. The results of investigations indicate that although interference from a students' first language is the major predictor of phonological errors (as most experienced EFL teachers already know), interference errors are only one of the types of errors found in the syntax, morphology and lexicon of student speech and writing in the target language (Richards 1971, Politzer 1974, Ervin-Tripp 1970, George 1972, and Grauberg 1971, as cited in Burt, 1975). For example, Grauberg (1971), found that "mother tongue interference could account for only 25% of the lexical errors, 10% of the syntactic errors, and none of the morphological errors in his students' essays" (Burt, 1975, p. 54).
Most of the current work in error analysis focuses on either the linguistic classification of errors or on the causes of errors made by adults learning English and other foreign languages. The primary causes of errors are:
Interlingual/Transfer errors: those attributed to the native language (NL). There are interlingual errors when the learner's L1 habits (patterns, systems or rules) interfere or prevent him/her, to some extent, from acquiring the patterns and rules of the second language (Corder, 1971). Interference (negative transfer) is the negative influence of the mother language (L1) on the performance of the target language learner (L2) (Lado as cited in Khodabandeh, 2007, p.8 ).
Intralingual/Developmental errors: those due to the language being learned (TL), independent of the native language. According to Richards (1970) they are items produced by the learner which reflect not the structure of the mother tongue, but generalizations based on partial exposure to the target language. The learner, in this case, tries to "derive the rules behind the data to which he/she has been exposed, and may develop hypotheses that correspond neither to the mother tongue nor to the target language" (Richards as cited in Khodabandeh, 2007, p.8 ).
A division of learners' errors according to whether they appear to originate from an L1 structure or L2-dependent rule construction has formed the mainstay of studies of L2 learner-language for over a decade. The computation of the relative frequencies of errors ascribed to these categories (occasionally in relation to differing levels of proficiency) has served a variety of research questions and theoretical issues. Perhaps the most important issue was whether or not L2 acquisition draws on those cognitive mechanisms and language processing strategies that have been identified for L1 acquisition. The significant number of target language-specific developmental errors found in the L2 speech of learners of different Lls and age groups played a major role in the reemphasis of Contrastive Analysis in L2 research and cleared the way for a research paradigm which focused on the commonalities at the product and process level (Zobl, 1980).
Zobl, (1980) demonstrated that both types of errors have their origins in the processing of properties of L2 input, and that errors which seem to show influence from the L1 presumably begin as L2-dependent developmental errors which are subsequently reinforced by an L1 structure compatible with the developmental error. Dulay and Burt (1972) have demonstrated the traditional distinction of developmental vs. transfer errors in table 1.
Table 1Distinguishing Features of Developmental and Transfer ErrorsProduct Level Developmental Transfer Errors are identical to those made by native learners (Mougeon and Hebrard 1975) Interference yields different types of errors than those that are de-velopmental (Dulay and Burt 1974a) Errors result from the creation of rules similar to those of native learners (Dulay and Burt 1974a)Errors reflect learner's use of L1 as a crutch at low level of L2 proficiency (Taylor 1974)Errors reflect learner's competence at a particular developmental stage (Richards 1971a)Errors reflect use of L1 as a source of hypotheses about L2 (e.g., Cancino, Rosansky and Schumann 1975 )Errors illustrate learner's attempt to build up hypotheses about the L2 (Richards 1971a)Process Level Origins of errors are within the L2 ( Richards 1971a)Errors reflect inability to separate the two languages (Richards 1971a) Errors reflect general characteris-tics of rule learning and language acquisition (Richards 1971a)Errors result from L1 habits (out-dated Contrastive Analysis posi-tion) Main strategies are simplification, generalization and reduction of grammatical redundancy (Dulay and Burt 1972)Errors represent an interlingual generalization (TayIor 1974)
Excerpted from Zobl (1980)
There is also another wildly known division of errors, global vs. local errors. Burt, (1975) discusses four aspects of English grammar that often cause global errors: basic word order, sentence connectors, psychological predicate constructions and selectional restrictions on certain types of verbs in sentential complements. The two last aspects share an important characteristic: they are exceptions to pervasive principles of English. In natural conversations, speakers constantly generate new sentences by applying the rules of the language they are speaking. Thus, even if a student has never heard a particular sentence, he can generate it if he has internalized (learned) the relevant grammar. However, in certain instances, English requires its own basic rules to be violated, and an unsuspecting student of English will apply a rule he has learned and consequently make an error. These kinds of errors, which are made by students from diverse language backgrounds, make it clear that the mother tongue is not the source of these errors. Rather, English itself is the 'culprit'(Burt, 1975). "Research has shown that accuracy of language use may be significantly improved with the mastery of the correct usage of the vocabulary item. It is in this sense that Gairns [and] Redman (1986) speak of the 'grammar of vocabulary' (Li & Chan, 1999, p. 85)". Examples of this include misuse of verb transitivity (*to discuss about something), confusion between a noun and verb (*to emphasise on something), preposition (*to walk under the sun), and so forth.
This grammar of vocabulary is what we are investigating in this paper. In the present study we focus on verb selectional restriction within subcategorization frame. The following questions will be answered:
Do subcategorizational differences between Persian and English lead to inaccuracy in production of English as L2 in institutes?
In case they lead to errors, do they change to correct form as the proficiency level increases?
A group of seventy learners, forty at elementary level and thirty at intermediate level, will participate in this study. They will be selected from different language institutes in Mashhad in order to remove any probable effect of familiarity with a specific verb use due to instruction of a specific course book. Participants are female students who have studied English at least for a year.
They will be given a list of 20 sentences; including verbs commonly used erroneously by Persian learners of English, and were asked to translate them into English.
The verbs selected for the study were based on hints in Common Mistakes in English and also based on teachers' experience about verbs EFL learners often make mistake about. The translations of verbs, which are used in Persian sentences, will be given to students on a piece of paper with a blank in front of them for students to write the English sentences. In order to make subjects produce the desirable sentences the verb to be used is given in parenthesis at the end of each Persian sentence. The sample paper to be used for data collection is available in appendix 1.
Procedure and Analysis
Papers will be distributed by the researcher. Learners will be given twenty minutes to write sentences. The same list of sentences will be distributed among intermediate learners to find out if the same kind of errors occurs in learners in higher level of proficiency. The major category of errors in this study is syntactic one. Sentences containing errors of subcategorization will be selected and other kinds of errors as well as correct sentences will be ignored. For example in the sentence *We should obey from teacher. only wrong usage of preposition is considered as error for the study and neglect of article for teacher will be ignored.