Introduction

Error Analysis (EA) has been studied since Corder introduced it in 1967. In “The Significance of Learners’ Errors,” Corder (1967) focuses on errors made by both children acquiring L1 and students learning L2. He agrees that with both L1 and L2, learners utilize a “definite system of language” (n.p.) throughout their language development. He emphasizes the fact that like L1 speakers,L2 learners make mistakes when speaking and writing, and mistakes are often corrected when learners notice them. In Corder’s (1967) perspective, errors are of two types: those “which are the product of such chance circumstances and those which reveal his [the learner’s] underlying knowledge of the language to date, or . . . his [the learner’s]  transitional competence” (italicized by the author) (p. 166). According to Corder (1967), errors of performance are called mistakes and systematic errors are defined as errors which refer to learners’ transitional competence. Corder (1967) questions the systems of L2 in relation to those of L1: “Are the systems of the new language the same or different from those of the language I know? And if different, what is their nature?” (p. 168). He agrees that not all but most errors made by L2 learners are caused by L1 interference.  Based on Corder’s (1967) work, researchers from different countries  (e.g., Bennui, 2008; Darus & Subramaniam, 2009;  Elkılıç, Han, & Aydin, 2009;  Kırkgöz, 2010; Lee, 2004; Masangya & Lozada, 2009; Mousavi &  Kashefian-Naeeini, 2011)  have focused their investigations on what types of errors students make when they write in English (L2). All these researchers found that the most common writing errors were grammatical errors. Bennui (2008), Darus and Subramaniam (2009), Kırkgöz (2010), and Lee (2004) also found lexical errors in students’ writings.

Some errors made by the participants of the studies above were a result of L1 interference. Interference can be defined as “the native language effect” (Brown, 1994, p. 26). When learners start studying a new language, their first language (L1) helps to facilitate the learning process of the new language (L2). At the beginning stages of L2 learning, the only reference learners have is L1, so they assume that L2 is similar to L1, according to Brown (1994). Brown (1994) also discusses interlanguage when he refers to the developmental process learners undergo to become competent in L2. He explains this process as “a systematic or quasi-systematic developmental process” (p. 27). Other researchers such as Kellerman and Sharwood, Selinker, and James refer to Brown’s (1994) “the native language effect” and “interlanguage” (pp. 26-27) as “cross-linguistic influence,” “language transfer,” and “transfer analysis” respectively (as cited in Alamin & Ahmed, 2012, p. 3). James explains the causes of errors as being “interlingual interference” and “intralingual interference” (as cited in Alamin & Ahmed, 2012, p. 3). “Interlingual interference” is the same as “the native language effect” (Brown, 1994). On the other hand, “intralingual interference” refers to errors caused by L2. In a book published in 2000, Brown refers to errors as interlingual and intralingual, too, and explains that intralingual errors refer to errors originated from “the target language, context of learning and communication strategies” (p. 218).

Literature Review

Type of Errors

Errors have been described by Burt and Kiparsky (as cited in Hendrickson, 1976) as being of two types:  local and global errors. Ellis (1994) explains that while local errors “affect only a single constituent in the sentence (for example, the verb), and are, perhaps, less likely to create any processing problems,” global errors “violate the overall structure of a sentence and for this reason may make it difficult to process” (p. 20).

In the literature, studies have focused on local and global errors. Lee’s (2004) study investigated how instructors corrected errors in students’ papers. The researcher found that there were a total of 19 types of errors in students’ papers and most were local errors. The majority of the errors were in noun ending, spelling, punctuation, verb tense, and article. Eight of the errors were lexical errors. Darus and Subramaniam (2009) found that errors in participants’ writings were in word choice, word order, subject-verb agreement, verb tense, prepositions, and singular/plural forms. Elkılıç, Han, and Aydın’s (2009) study, in contrast, analyzed papers for punctuation and capitalization to find whether punctuation and capitalization in English writing caused problems and whether as students became more fluent in L2 the problems decreased. The results showed that“both intermediate and upper-intermediate students committed fewer interference errors . . . than general errors” and that as students became more fluent in L2, errors in punctuation and capitalization reduced (p. 1). Masangya and Lozada’s (2009) investigation was somehow similar to Elkılıç, Han, and Aydın’s (2009) in terms of the relationship between exposure to L2 and errors in English essays. Their findings showed that students who had more exposure to L2 had fewer errors than those who had less exposure, and the errors were in wrong case, fragmentation, parallelism, punctuation, and verb tense. Also, Mousavi and Kashefian-Naeeini’s (2011) study analyzed errors made by graduate Iranian students. They found that the errors were in orthography and grammar. Additionally, James (1988) listed the most common writing errors made by ESL learners:  tenses, prepositions, and poor vocabulary.

Cause of Errors

Errors happen for different reasons. According to Brown (1994), some errors are a result of interlingual interference or L1 interference. As beginning learners have not learned much about L2 yet, they assume that it “. . . operates like the native language” (Brown, 1994, p. 65). Bennui  (2008) analyzed and described aspects of L1 interference in students’ paragraph writing, and found problems  with lexicon since students translated words from L1 (Thai) to L2 (English). Problems with word order, subject-verb agreement, verb tense, prepositions, and noun determiners, all due to L1 syntactic interference, were also found.  Bennui (2008) discovered errors caused by L1 discourse interference, too.  Darus and Subramaniam (2009) noticed that L1 also caused interference. Students assumed that if L1 did not have a certain rule, L2 did not have it either, such as –s for countable plural nouns (regular forms). In Darus and Subramaniam’s (2009) participants L1 (Malay), “there is no plural marker for a noun” (p. 492). What is more, Kırkgöz’ (2010) study analyzed beginning students’ essays for punctuation and capitalization, and found that their errors were due to L1 (Turkish) interference. Falhasiri, Tavakoli, Hasiri, and Mohammadzadeh (2011) investigated frequency of errors in participants’ writings and effectiveness of feedback on error reduction. They concluded that the most frequent errors resulted from L1 (Persian) interference, and misuse of prepositions was the most frequent error of interference. The most frequent error was participants’ omission of –s in the plural form of countable nouns.  The second and third most frequent errors were omission of articles and omission of –s in the third person singular of verbs in the simple present tense.

Intralingual errors were also identified in Kırkgöz’ study (2010). Overgeneralization was the type of intralingual error cited by her.  By overgeneralization she meant “negative transfer of language items and grammatical rules in the target language, incomplete application of rules” (p. 4356).

In contrast, Mousavi and Kashefian-Naeeini’s (2011) study presented several other different causes of learners’ errors based on results of surveys given to participants. Some participants blamed their problems on their lack of practice writing in English; others attributed their problems to their Iranian instructors’ lack of experience as teachers; still some others criticized the environment for their lack of motivation. Moreover, in interviews conducted with Arab students about causes of problems with the English language, Al-Khasawneh (2010) concluded that participants’ problems were attributed to “their weak foundation, environment, and methods of teaching English in their countries” (p. 16).

The discussion above indicates that EFL writing errors identified in studies are mostly local errors and usually caused by L1 interference.  The studies report findings in their contexts.  Each study increases our understanding of this feature in EFL writing.  In light of this discussion, the goal of the current study was to add to this collective knowledge being built by investigating in this specific context what type of errors participants made in their essays and to try to identify why they made the specific errors. Thus, the following research questions guided the study:

RQ1 What type of errors (local and/or global) do the Turkish students in the study make in their English essays?

RQ2 Why do students make these errors?

Methodology

Participants

In this study, the participants were 17 freshman Turkish students (sixteen males and one female) enrolled for the first time in the first writing course at Eskişehir Osmangazi University Electrical/Electronic Engineering (EEE) Department, Turkey, in the fall 2011. Participants’ mean age was 20.1 years. In addition to studying English in junior and senior high schools, all participants attended the university Foreign Languages Department Preparatory English Program for one year before they entered the EEE Department. Thus, by the time they entered the EEE Department, their level of fluency was very similar; that is to say, they were all at the intermediate level of fluency in English.

Data Collection

The data source for the study was a comparison/contrast essay (Appendix A). For the final examination, participants were instructed to write a 4-paragraph essay only (an introductory paragraph, two body paragraphs, and a concluding paragraph) on one of the five possible topics provided in the exam, an essay they could complete within the 90 time limit.  Approximately 80% of the  students chose the second topic, Two cities; about 15% of the students, the third topic, Life in high school and life at the university; and about 5% of the  students, the first topic, Two professions – engineering and teaching.17 essays were analyzed.  Corder (1974) was used primarily as the data analyses guide, i.e., sample collection, error identification, error description, error explanation, and error evaluation.

Data Analysis

To respond to the study research questions, the study identified errors in lexicon, grammar, and syntax. To be specific in the description of errors, the nine most prevalent errors were in word choice, article (addition and omission), preposition (addition, omission, and misuse), punctuation (omission and misuse), singular/plural noun agreement, spelling, verb tense, sentence fragment, and subject-verb agreement. Lexical errors were the most serious errors, errors that often prevented the researcher from understanding what the participants wrote. Table 1 displays type of errors, number of errors, percentage, and mean value of errors made by the 17 participants in 17 comparison/contrast essays:

The errors were classified into two groups: global and local errors (Burt & Kiparsky, as cited in Hendrickson, 1976) (explained above). In the present study, most errors in word choice refer to global errors because most caused lack of understanding (See Table 2 for examples of global and local errors.). Instructors should always remember that both global and local errors are part of learning L2. As Brown (1994) emphasizes, “Errors are . . . windows to a learner’s internalized understanding of the second [and foreign] language[s], and therefore, they give . . . teachers something observable to react to” (p. 27).

Findings and Discussion

Table 1 below shows the nine most prevalent errors made by the participants, number of errors, percentage of errors, and mean value of errors: word choice = 11.17; article (addition and omission) = 3.82; preposition (addition, omission, and misuse) = 2.47; punctuation (omission and misuse) = 2.05; singular/plural noun agreement = 1.76; spelling = 1.70; verb tense = 0.70; sentence fragment = 0.64; and subject-verb agreement = 0.47. As seen in Table 1, Word choice was by far the most common and serious error, the error that resulted in frequent misunderstanding.

Table 1: Error identification and description

Table 2 below displays some examples of the most and least common errors found in the 17 essays.

Table 2: Type and example of errors

Table 2 above shows at least one example of each type of error made by the participants. Word choice which was the error that caused more problems for the participants than other errors and led to misunderstanding was sometimes totally incomprehensible. A few times the researcher could figure out what the participant tried to say but not all the time. The results of the present study support those of Darus and Subramaniam’s (2009) which confirmed participants lack of proper words, but theirs, as seen in the examples, did not appear to cause problems with comprehension. The same was true for some of Kırkgöz’ (2010) examples taken from participants’ essays. The words were translated from L1 (Turkish) to L2, but they did not prevent comprehension since Kırkgöz was a speaker of L1. Had their instructor been a NS of English, the choice of words might have resulted in misunderstanding. Also, the current study supports Rabab’ah’s (2005) study. Rabab’ah (2005) indicated that based on his experience teaching undergraduate and graduate English majors, Jordanian students “lack[ed] the necessary vocabulary they need[ed] to get their meaning across” (p. 183).

In reference to definite and indefinite articles, the Turkish language has neither; therefore, sometimes participants used articles when they did not need them or did not use them when necessary. Kırkgöz’ (2010) study using Turkish learners as participants is supported by the current study. In her study, Kırkgöz’ participants added, omitted, and misused both definite and indefinite articles due to lack of articles in the Turkish language, according to Kırkgöz. She quoted Pascasio and Stockwell to explain why it is difficult for learners to learn a grammar rule that belongs to L2 but not to L1.

For most L2 learners, prepositions may be difficult to learn, too. In Turkish, there is one suffix which is equivalent to the prepositions at, in, and on in English. The specific suffix is spelled differently in various situations so as to harmonize with the sound that comes before the suffix, for example, ev (house/home) and evde (in the house or at home); araba (car) and arabada (in the car); mutfak (kitchen) and mutfakta (in the kitchen); tuvalet (restroom) and tuvalette (in the restroom). As a result, when Turkish speakers learn English prepositions such as at, in, and on, they become confused and do not know when to use them. In the present study, sometimes participants either omitted prepositions altogether or added wrong ones, as seen in two examples given in Table 2. In the sentence “Most people work on there,” on there is the translation for the word orada (da is a suffix translated as on in this example). The current study confirms what was found in Darus and Subramaniam’s (2009) investigation. Kırkgöz’ (2010) also found errors in the use of prepositions in participants’ essays and one specific error was the same type that the researcher of the present paper found: addition of a preposition. The example that Kırkgöz (2010) gave was the following: “Suzanne is on downstairs” (p. 4355). Since in Turkish a suffix is added to the word downstairs which becomes aşağıda (da is the suffix equivalent to at, in, and on), the participant may have thought in L1 and translated into L2, and the result was the added preposition to the English sentence. Masangya and Lozada’s (2009) results are also supported by the present study. The researchers confirmed how difficult it was for learners to learn L2 prepositions. They found that the highest number of errors committed by their participants was in the use of prepositions which accounted for forty-six percent of the total. Blake’s study (as cited in Masangya & Lozada, 2009) had found the same type of error and the explanation given by Blake was that Filipino students found it hard to learn the usage of prepositions in English since in their L1 there was only one preposition, sa. Therefore, Masangya and Lozada (2009) blamed learners’ confusion on the number of prepositions in the English language.

In the current study, punctuation was another case of L1 interference; nevertheless, punctuation did not cause many problems. The errors might have been due to participants’ lack of attention when they were taught punctuation rules. It may also be possible that participants were never taught punctuation rules or punctuation rules were simply somehow difficult for them to learn. The present study findings support Elkılıç, Han, and Aydın’s (2009). In their investigation on capitalization and punctuation errors of Turkish participants, the percentage of errors in punctuation was not given, but it appeared that the Turkish participants did not have significant problems with punctuation. The conclusion that Elkılıç, Han, and Aydın’s (2009) came to was that students had a tendency to resort to L1 to punctuate sentences in L2. In contrast, Mousavi and Kashefian-Naeeini’s findings (2011) showed that Malay participants had more errors in the use of punctuation than in any other area as far as basic skills were concerned. Mousavi and Kashefian-Naeeini (2011) affirmed: “According to the results of the last item [‘using punctuation accurately’], only 2.4% of the respondents do not have any problem...using punctuation accurately...and 30.1% have less difficulty, whereas 39.8% have difficulty and 27% much difficulty. . .” (p. 600).

Additionally, in this study, singular/plural noun agreement errors in participants’ essays were apparently caused by L1 interference. In the Turkish language, a noun that follows phrases/words, such as a lot of, a number of, many, and any number greater than one, does not require the plural marker. Thus, participants had many errors in noun agreement when they used the phrases/words above, as shown in the examples displayed in Table 2. The present study supports Kırkgöz’ (2010) findings. In her study, she, a native speaker of Turkish, gave examples of the Turkish rule and participants’ errors: “He is drinking three cup of coffee. We have a big garden and three dog. There is [sic] two telephone on the table. There is [sic] a few apple in the basket” (p. 4355).

Furthermore, only seven percent of the errors in participants’ essays were in spelling. The researcher has observed in many years that she has taught English to Turkish students that spelling tends not to cause problems for them. The reason may be that the pronunciation and spelling of English words differ so much that Turkish students pay very close attention to English spelling. The present study supports Kırkgöz’ (2010) study. Only eight percent of the errors found in her participants’ essays related to spelling. In contrast, Mousavi and Kashefian-Naeeini’s (2011) study of Iranian participants found that most participants had difficulties with spelling. The researchers divided the participants into four groups to explain the percentage of errors in spelling: those that had much difficulty with spelling (17.9%); some difficulty (28.6%); not so much difficulty (36.9%); and no difficulty (16.7%).

Also, verb tense was not a very frequent error in participants’ essays. In Turkish and in English, the tenses are somehow similar. Sometimes, though, they are not used the same way in the two languages, and that may be the reason problems occur. In the Turkish language, the verb tenses are the following: present, present progressive, future, definite past, indefinite past, necessity, subjunctive, conditional, and imperative (Halman, 1981). It appears that the only verb tense the Turkish language does not have when compared to the English language is the present perfecttense.The example displayed in Table 2 may be a problem of confusion of tenses in the mind of the participant or the fact that Turkish does not have the present perfect tense. While an English speaker says, “I have lived in Mersin for 20 years,” a Turkish speaker says, “I am living in Mersin for 20 years.” On the other hand, Bennui’s (2008) participants from Thailand said that in Thai, there was no time and tense relationship, that is, time indicating what tense the speaker or writer was expected to use. The verb form remained the same no matter the time-- present, past, future, etc. One of Bennui’s (2008) participants wrote: “First, it makes me love (to) read when I was young until future I read books.” Bennui (2008) explained that the participant used the simple present tense in a sentence that was in the past. The participant also used the future word “to express a future in the past” (p. 85).

It is important to note that the participants who took part in the current study learned what was taught in the course about the development and organization of comparison/contrast essay. Their introductory paragraph, body paragraphs, and concluding paragraph were developed and organized properly, and were coherent and unified. Very few errors in sentence fragments were detected, as indicated in Table 2. Perhaps, the reason was that sentence fragments were taught in the course. Darus and Subramaniam (2009) also found errors in sentence structure in their participants essays, and the errors resulted in sentence fragments, such as “Secondly, the dirties on the kitchen area” (p. 493).

Additionally, few errors in subject-verb agreement (simple present tense) were seen in participants’ essays. It is relevant to mention that subject-verb agreement errors were not caused by L1 interference. For those who made this type of error, it may be possible to explain the errors as a result of carelessness or lack of practice communicating in both written and spoken English. In Turkish, verbs are conjugated; therefore, for each personal pronoun, verbs have different endings when conjugated, no matter the tense. Bennui (2008) had this to say about the Thai language and the reason participants in his study made errors: “The agreement of the subject and verb in relation to the tenses containing many sentences is not found in Thai structure such as ‘Everybody have different things,’ and ‘My friend usually know my books’” (p.85).

Conclusion

The study was set up to analyze errors in 17 English essays written by 17 Turkish participants in a freshman composition course. The study followed Corder’s (1974) five steps: sample collection, error identification, error description, error explanation, and error evaluation. The errors identified were in lexicon, grammar, and syntax. The errors were classified into global and local errors, global errors meaning those that prevented understanding and local, those that did not cause lack of comprehension but were awkward. Word choice caused most of the global errors. The researcher hopes that more studies that focus on the analysis of lexical errors in L2 writing be conducted since there are very few (Llach, 2005; Mahmoud, 2011). The more studies conducted, the more instructors will learn about the difficulties L2 learners face when they lack depth and breadth of vocabulary. Instructors will then be better prepared to help L2 learners to avoid global errors.

Pedagogical Implications

This study has a number of pedagogical implications particularly related to writing errors and teaching interventions. Because of the relative importance of global versus local errors instructors should emphasize the correction of global and not local errors. If errors do not prevent comprehension, instructors should not bother with them. What is important is that learners communicate their thoughts in an understandable manner (Hendrickson, 1976). Brown (1994) believes that the solution to errors of interference is to make students aware of the origin of the error. Brown (1994) says that good students will get rid of the interference once they understand where it originates. For those students who are unable to do so, he suggests that “thinking directly in the target language usually helps to minimize interference errors” (p. 27).

In addition to engaging in EA research, teachers should be aware of L1 language conventions and know how these are manifested in issues of language interference. Darus and Subramaniam (2009) believe that “teachers who can analyze and treat errors effectively are better equipped to help their students become more aware of their errors” (p. 486). They continue by saying that “EA is the best tool for describing and explaining errors made by speakers of other languages” (pp. 486-87). In addition, Al-Khasawneh’s (2010) participants suggest that for them to improve their writing skills, their university, the learners themselves, the instructors, and the writing process should all be taken into consideration. Mansangya and Lozada (2009) conclude after analyzing participants’ essays that the more exposure learners have to L2, the fewer errors they make. As for punctuation and spelling, in particular, Elkılıç, Han, and Aydın (2009) suggest that L1 and L2 punctuation should be compared, and punctuation in L2 should be taught straightforward. In addition, Masangya and Lozada (2009) think that for learners to learn how to spell in L2, they have to be exposed to as many words as possible. Corder (1967) affirms: “Making a learner try to discover the right form could often be more instructive to both learner and teacher” (p. 168). His perspective presented here has been borrowed from Carroll.

Study Limitations

The researcher cannot deny that the current study has many limitations. The first one is the small number of participants. The second one is the small number of essays analyzed. It is possible that with a much larger number of participants and essays, results might have been different and varied. Third, had the same group of participants been given a chance to write a second comparison/contrast essay with a list of words on the topic assigned, the researcher would have compared essays written by the same group of participants and would have found different results. Fourth, it is also possible that since the 17 essays were graded by the same instructor, findings would have been different had another instructor graded the same essays.

References

Alamin, A., & Ahmed, A. (2012). Syntactical and punctuation errors: An analysis of technical writing of university students Science College, Taif University, KSA. English Language Teaching, 5(5), 2-8.

Al-Khasawneh, F. M. S. (2010). Writing for academic purposes: Problems faced by Arab postgraduate students of the College of Business, UUM. ESP World, 2(28), 1-23.

Bennui, P. (2008). A study of L1 interference in the writing of Thai EFL students. Malaysian Journal of ELT Research, 4, 72-102.

Brown, H. D. (1994). Teaching by Principles: An Interactive Approach to Language Pedagogy. Englewood Cliffs, NJ: Prentice Hall Regents.

Brown, H. D. (2000). Principles of Language Learning and Teaching. England: Longman.

Corder, S. P. (1967). The significance of learners’ errors. IRAL 4, 161-170. ERIC Document Reproduction Service, ED 019903.

Corder, S. P. (1974). Error Analysis. Oxford: Oxford University Press.

Darus, S., & Subramaniam, K. (2009). Error analysis of the written English essays of secondary students in Malaysia: A case study. European Journal of Social Sciences, 8(3), 483-495.

Elkılıç, G., Han, Turgay, & Aydın, S. (2009). Punctuation and capitalisation errors of Turkish EFL students in composition classes: An evidence of L1 interference. In 1st International Symposium on Sustainable Development, June 9-10, 2009, Sarajevo, Bosnia and Herzegovina, 279-284.

Ellis, R. (1994). The Study of Second Language Acquisition. Oxford: Oxford University Press.

Falhasiri, M., Tavakoli, M, Hasiri, F., & Mohammadzadeh, A. R. (2011). The effectiveness of explicit and implicit corrective feedback on interlanguage and intralingual errors: A case of error analysis of students’ compositions.    English Language Teaching 4(3), 252-264.

Halman, T. S. (1981). 201 Turkish Verbs Fully Conjugated in All the Tenses. New York: Barron’s Educational Series, Inc.

Hendrickson, J. M. (1976). Error analysis and selective correction in the adult ESL classroom: An experiment. ERIC Document Reproduction Service, ED 135260, 1-21.

James, C. (1988). Errors in Language Learning and Use: Exploring Error Analysis. Harlow,Essex: Addison Wesley Longman Limited.

Kırkgöz, Y. (2010). An analysis of written errors of Turkish adult learners of English. Procedia Social and Behavioral Sciences, 2, 4352-4358.

Llach, M. P. A. (2005). The relationship of lexical error and their types to the quality of ESL compositions: An empirical study. Porta Linguarum, 3, 45-57.

Lee, I. (2004). Error correction in L2 secondary writing classrooms: The case of Hong Kong. Journal of Second Language Writing, 13(4), 285-312.

Mahmoud, A. (2010). The role of interlingual and intralingual transfer in learner-centered EFL vocabulary instruction. Arab World English Journal, 2(3), 28-49.

Masangya, E. M., & Lozada, L. (2009). An investigation on the relationship between the language exposures and errors in English essays of high school students. Philippine ESL Journal, 2, 31-50.

Mousavi, H. S., & Kashefian-Naeeini, S. (2011). Academic writing problems of Iranian post-graduate students at National University of Malaysia (UKM). European Journal of Social Sciences, 23(4), 593-603.

Rabab’ah, G. (2005). Communication problems facing Arab learners of English. Journal of Language and Learning, 3(1), 180-197.