Comparing Two Feedback Methods on Iranian EFL Learners’ Translation

Details

Written by Siamak Mazloomi

Abstract

The current study attempts to compare the effects of two feedback methods, recast and metalinguistic feedback, on Iranian EFL learners’ translations.  Forty subjects were homogenized and received eight sessions of treatment and identical pretests and post-tests in two groups, each receiving the same texts and tests in order to identify the effect of the received feedback on their translation of complex-compound structures. The results obtained from the T-tests indicate that both methods contribute to the development of the EFL learners’ translation skills (sig. = .018, p< .05 and sig. = .000, p<.01); however, the values obtained by the group who received metalinguistic feedback demonstrate that metalinguistic feedback seems to be a more effective way to treat the EFL learners’ errors in their translations.  

Keywords: Corrective feedback, Translation tasks, Recast, Metalinguistic feedback

1. Introduction: Corrective feedback and its typology

Corrective feedback has been defined as “any reaction of the teacher which clearly transforms, disapprovingly refers to, or demands improvement of the learner utterance” (Chaudron 1977, as cited in Mackey et al. 2007).  Corrective feedback can be overtly corrective (explicit) or may not overtly mark the learner’s production (implicit). (Mackey et al. 2007) Explicit corrective feedback provides learners with a meta-linguistic explanation or overt error correction. On the other hand, implicit corrective feedback indirectly and incidentally informs learners of their errors. In this case, since the correction is unobtrusively provided so as not to interrupt the flow of the conversation, ungrammaticality is expected to be inferred. Recasts,confirmation checks, clarification requests, repetitions, and even paralinguistic signs such as facial expressions can all constitute implicit corrective feedback (Long 1996).

Lightbown and Spada (2006: 125-128) provide us with a typology of various feedback applied in the classroom environment by the ESL/EFL teachers. Their categorization includes explicit correction, recasts, clarification requests, metalinguistic feedbacks, elicitations, and repetitions as different types of feedback utilized by teachers.  Hernandez et al. (2010: 254) reports their field study, “The type of correction mostly used by teachers is teacher and peer correction. The strategies more frequently used are recasting, clarification requests, elicitation, interruption and body language. On the other hand, explicit correction seems neglected as well as metalinguistic awareness.”

The current article focuses on translation as a means of SLA instruction (Machida 2008) as well as an independent field of study; only two of the above-mentioned categories, which seem the most popular and fruitful, are investigated for a translation task in EFL classrooms.  The first one is the most usual in translation instruction, recast or namely repetition of the students’ erroneous utterances. (Lightbown & Spada 2006) And the less explicit one is metalinguistic feedback, which “contains comments, information, or questions related to the correctness of the student’s utterance, without explicitly providing the correct form.” (Lightbown & Spada 2006: 126) Here, by recast as feedback on the learners’ translation, I mean the repetition of their erroneous translation in a well-formed target translation. And by metalinguistic feedback, I mean giving some hints, comments and necessary information on their erroneous translations in order to help them come up with some correct versions of their translations.       

The subjects in the current study are translation trainees and have to attend their translation classes. Thus, translation is being practiced in their classes as the necessary skill to satisfy the objectives of their courses. They are to hand in their translations of the distributed texts every session. These translations are then corrected by the teacher in two ways: explicitly through recasts and implicitly through metalinguistic feedback.

1.1 Is corrective feedback helpful?

Whether or not corrective feedback is helpful in L2 learning has been examined in both laboratory and classroom settings. The results from laboratory studies have demon­strated a facilitative effect for corrective feedback on L2 development. Also, studies conducted in classroom settings have generally been supportive of the claim that corrective feedback positively affects learning. Moreover, there is a current concern among teachers and researchers that the absence of explicit instruction and correction will lead to early fossilization of errors (Lightbown & Spada 1999).  As a result, corrective feedback is now considered to be helpful for L2 learning by most researchers in the field of SLA, although different types of feedback need to be studied independently and together for information about their role in learning (Mackey et al. 2007).

            Data from French Immersion classrooms also provide an explanation of why learners need to attend to meaning as well as linguistic aspects which otherwise may go unnoticed, unprocessed, and unlearned (Mackey et al. 2007). In so doing, translation tends to be one the most efficient and essential ways to clarify meaning. On the other hand, in his Noticing Hypothesis, Schmidt (1990, 1995, 2001) argues that noticing is requisite for learning, stating that learners must consciously pay attention to or notice input in order for L2 learning to proceed. Those subscribing to the Noticing Hypothesis and the Interactionist model also recognize the value of corrective feedback, assigning it a facilitative role in drawing learner attention to form (Long 1996). From this perspective, corrective feedback serves as a stimulus for noticing because such feedback triggers learners to recognize the gap between their interlanguage and the target norm – this process in turn leads to subsequent grammatical restructuring.  Translation as a means of meaning clarification can be a type of noticing and awareness raising. 

1.2. Translation as a teaching method

Malmkjær (1998: 6) counts the arguments made against the use of translation in EFL/ESL settings, based on Berlitz (1907), Lado (1964) and Gatenby (1967): First, translation is basically different from those four skills involved in language competence: reading, writing, speaking and listening, and it would take time to learn, which could have been spent learning those major skills. Translation is also unnatural, misleading students into thinking that there is one-to-one correspondence between various items in languages. Further, translation encourages students to hold on to their native language habits, causing interference and interrupting thinking in the language being learnt. Translation, moreover, does not provide an appropriate way to test language skills because you cannot compose freely and naturally in L2 if L1 is constantly there in the form of an ST. In fact, translation exercises are thought to be appropriate only for the translation teaching classroom.

            To challenge such arguments, Carreres (2006) conducted a survey on thirty one Spanish language students at the University of Cambridge. Over half of the students (54%) agreed that translation was a more effective method than other methods, although it was not uniformly popular among them, and it should be taught as part of a modern languages undergraduate degree. Källkvist (2008) also undertook a longitudinal study over thirteen weeks “in which two groups of randomly assigned advanced-level [Swedish] learners of English were given two different types of form-focused exercises, only one of which involved translation” (Källkvist 2008: 183). Källkvist concludes that (2008: 199):

If we expect and aim for our learners to be able to use the L2 well when communicating in situations in which they are required to translate and in situations in which they need to express themselves directly in L2, it seems fully reasonable that we provide them with exercises and rich, varied, and enhanced input of either kind.

In spite of such arguments, in recent years, the necessity for integration of explicit instruction into communicative approaches has become obvious. In this view, language is not only understood as a communication instrument, but also as a reflection of the context in which it is used, such as the speakers’ L1 culture, and the sociolinguistic features of the situation. On the other hand, Constructivism, the concept currently pervasive in the educational field, sees that individual learners construct knowledge for themselves, using their prior experience and the range of contextual elements they perceive. Such learning is also seen as a personal process. (Machida 2008) A Constructivist approach to learning emphasizes the provision of authentic, challenging projects for the learners. Projects which are meaningful to the learners, i.e. ones in which they can incorporate their experience outside of the classroom, which are set in problem solving contexts, involving peers and teacher (expert) in the learning community, appear to promote effective learning. (Takimoto & Hashimoto 2010)

       Accordingly, one dramatic change in recent years has been the re-evaluation of L1 use in L2 instruction. In a turnaround from the Direct Method, the Audiolingual Method and Natural Approaches, it is now seen as potentially beneficial rather than damaging. According to our current understanding of vocabulary storage in the brain (e.g. the connectionist model), bilinguals access one common storage system containing both L1 and L2 vocabulary. L1 is thus considered to assist learners’ comprehension of L2 by creating more networks between nodes (ideational representation and words) in their long term memory.   (Machida 2008)  As a result, translation can be considered as a means of L2 instruction, especially in EFL classes. It provides the learners a challenging task to comprehend the source text in order to produce an authentic target one. Takimoto and Hashimoto (2010: 86) in their study and survey on Japanese college students have concluded, “ … students are engaged in active learning through interpreting and translation activities. Interpreting and translation encourage learners’ ‘intercultural exploration’, which in turn promotes a deeper understanding of both L1/C1 and L2/C2. In addition, interpreting and translation require learners to work within various constraints, which actually encourages intercultural language learning among students. Furthermore, the study indicates that the students felt that they acquired various skills including ‘cognitive skills’.”

       It is worth mentioning that the subjects in the present study are English translation students; thus, translation is not an irrelevant skill/art to them. Moreover, when they were asked if they found translation effective in their process of language learning, they all agreed that translation has affected and reinforced other language skills in them positively.

       The kind of translation tasks applied in the current study includes complex and compound-complex sentences (including adverb, adjective and noun clauses as well as their reduced forms) to be translated by the learners because in my expeience, these structures have been considered as the most problematic for the language learners and student translators. The learners’ translations in the current research have received two kinds of feedback especially on the above-mentioned structures:  recast and metalinguistic feedback.     

1.3.   Research questions

This study investigates three main questions regarding the effects of two different kinds of feedback on EFL learners’ translation competence.  The research questions are:

1. Do EFL learners’ translation skills significantly improve as a result of receiving recast as feedback on their translations?
2. Do EFL learners’ translation skills significantly improve after receiving metalinguistic feedback on their translations?
3. Is there any significant difference between the translation skills of the participants receiving recast as feedback and those receiving metalinguistic feedback?

In order to investigate the above-mentioned research questions empirically, the following null hypotheses have been tested:

1. There is no statistically significant difference between the pre-test and post-test scores of the participants who receive recast as feedback on their translation tasks.
2.  There is no statistically significant difference between the pre-test and post-test scores of the participants who receive metalinguistic feedback on their translation tasks.
3. There is no statistically significant difference between the post-test mean scores of the participants who receive recast as a feedback and those who receive metalinguistic feedback on their translation tasks.  
4. Method

2.1. Participants

The subjects in the current study were randomly selected from among the sophomore English Translation students in Islamshahr Azad University.  The type of sampling employed in this study is cluster sampling, randomly selecting groups of individuals rather than single individuals.  There were two 20-student classes all attending a course in General Text Translation by the same teacher.  These 40 subjects were selected and homogenized from among some 50, who had all attended a standardized TOEFL proficiency test, having a correlation of 0.87. The resulting sample is two 20-student groups of men and women at various ages ranging from 20 to 25.

2.2. Instrumentation

            The instruments applied in the current research were (a) a standardized TOEFL test (r = 0.89) to homogenize the participants, (b) two authentic one-paragraph texts (about 200 words) to be translated as the pre-test and post-test in each class. These two paragraphs as well as other paragraphs applied in the treatment are at the same level of difficulty and include complex or complex-compound sentences, which are the major concern of the present research.  

            The TOEFL test was administered in both classes prior to the beginning of the treatments to homogenize the two groups and measure their proficiency level. 

            The two one-paragraph texts applied as the pre-test and post-test were almost at the same length and readability.  The learners’ translations were rated by two independent raters, having a correlation of 0.74, which is moderate but quite acceptable for such subjective ratings.  The raters have both been translation instructors and translators for at least five years. The definite scores of the subjects on pretest and post-test translations were determined by calculating the mean of each student’s pairs of scores received from both raters. 

            It is worth mentioning that the raters were asked to consider only the sentences that included complex or compound-complex structures and score the students based on their translations of those sentences, ignoring their translations of simple sentences in the texts since the learners had received corrective feedback (recast and metalinguistic feedback) only on complex-compound and complex structures.

2.3. Procedure

The design of this study is the experimental pretest and post-test equivalent-groups design. A language proficiency test was applied in order to homogenize the participants in both groups.  The correlation between the two groups was 0.87, which demonstrates a high correlation and homogeneity between the two groups of students.

            Before the onset of the treatment in both groups, the subjects were asked to attend a pretest, including the translation of an authentic 200-word text into Persian, containing complex and compound-complex structures.  Their translations were rated by two independent raters, having a correlation of 0.74.  The raters have both been translation instructors and translators for at least 5 years.           

The two homogenized groups received 8 sessions of recasts as a kind of feedback and metalinguistic feedback on their translations.  In every session of the treatment, the students were provided with a short authentic text of about 200 words and ample time to translate in the classroom. These texts included complex and compound-complex sentences (some dependent and independent clauses as well as their reduced forms). In the first group, the teacher gave them recast feedback on their translations of the compound-complex sentences, presenting them with the correct forms of translations. In the second class, the subjects received metalinguistic feedback on their translations of the compound-complex sentences, provided with some comments and hints to help them arrive at more appropriate translations by themselves.  

            After the treatment, the subjects were asked to attend a post-test, including the translation of a 200-word text into Persian, including complex and compound-complex structures. Their translations were rated by two independent raters (r = 0.74). The subjects’ final scores both on the pre-test and the post-test were obtained by calculating the mean of both scores given by the raters on each test.

            In order to establish the significance of the difference between the pretests and the post-tests in both groups, a T-test was applied.  And in order to realize whether there is any statistically significant difference in the mean scores obtained from the post-tests in either group, another T-test was computed. 

Data analysis and results

The descriptive statistics obtained from the scores in pretest and post-test in each group can be seen in Table 1.

Table 1. Group Statistics

The low standard deviations and standard errors of measurement contribute to the reliability of our pretests and post-tests.  First, a T-test was calculated between the pretests and post-tests of the two groups, the first one receiving recast as feedback and the second one receiving metalinguistic feedback on their translations.  The results of this paired T-test are shown in Table 2.

Table 2.  Paired samples T-test of pretests and post-tests in both groups

The results of the T-test in the pairs in Table 2 indicate that in both groups, the treatment has produced a statistically significant difference between the mean scores obtained in the pretests and those achieved in the post-tests in each group, with the first treatment at 95% probability (p< .05) and the second treatment at 99% (p< .01).  This result demonstrates that in both groups, the kind of feedback provided on the learners’ translations was significantly influential; moreover, the same result reveals the fact that the second treatment, i. e. metalinguistic feedback, in the second group was more affective in improving the learners’ translation skills.  

            An independent T-test was also applied to the mean scores of the two pretests and post-tests in order to indicate which one of the treatments was more effective.  The results of this independent T-test are illustrated in Table 3.

Table 3. Independent samples T-tests between the the mean scores of the two pretests and post-tests

Since the Levene’s Test’s significance in both pretests and post-tests is more than .05 (.89 and .59, respectively), the variances in both should be assumed equal.  The T-value of the pretests (.83) indicates no significant difference between the mean scores of the tests before the treatments; however, the T-value of the post-tests (.013) reveals a significant difference obtained due to the different treatments (p< .05). As mentioned before, the second treatment (metalinguistic feedback) more significantly ensures the learners’ development in their translation skill due to its higher probability (Table 2).  Also the mean score obtained by the students in the second group (Table 1) definitely explains the stronger impact of the second treatment on achieving better results.

Conclusion

The results of the current study indicate that metalinguistic feedback is better feedback for translation courses than recast; that is, instructors and teachers do not need to repeat the whole translated sentence in order to correct their students; it is more helpful to provide comments, hints and information on grammatical clues, lexical points or even functional perspectives to help learners arrive at their own translations.  In fact, in a modern application of translation as a teaching method (e. g. Machida, 2008), translation is considered as a way of scaffolding the learners’ comprehension by negotiating the meaning of the written or spoken utterances in their native language.  The shift to the goal of meaning negotiation, according to Long (1996), is one of the outcomes of the modern Interaction Theory, which mediates the learners to improve through their zone of proximal development (ZPD) in a constructionist perspective. 

In a nutshell, translation is today considered more a means of better meaning negotiation rather than a goal, and teachers can help their pupils to achieve better skills in doing so by guiding them through metalinguistic feedback. 

References

Carreres, Angeles. 2006. “Strange bedfellows: Translation and language teaching. The

teaching of translation into L2 in modern languages degrees; uses and limitations”. Sixth Symposium on Translation, Terminology and Interpretation in Cuba and Canada: December 2006. Canadian Translators, Terminologists and Interpreters Council (online). Available at http://www.cttic.org/publications_06symposium.asp (accessed 22 April 2012).

Chaudron, Charles. 1977. “A descriptive model of discourse in the corrective treatment

of learners' errors”. Language Learning: 27, 29-46.

Hernández Méndez, Edith, Gómez Arguelles, Lizbeth & Jiménez Castro, Ana Bertha.

2010. “Oral corrective feedback: Some ways to go about it.” FEL International.

254-270.

Källkvist, Marie. 2008. “L1-L2 translation versus no translation: A longitudinal study of

focus-on-formS within a meaning-focused curriculum.” Lourdes Ortega and Heidi Byrnes eds. The Longitudinal Study of Advanced L2 Capacities. London and New York: Routledge. 182-202.

Lightbown, Patsy.M. & Spada, Nina. 1999. How languages are learned: Revised

edition. Oxford: Oxford University Press.

Lightbown, Patsy M. & Spada, Nina. 2006. How languages are learned. Oxford:

Oxford University Press.

Long, Michael H. 1996. “The role of linguistic environment in second language

acquisition.” William C. Ritchie & Tej K. Bhatia eds. Handbook of second language acquisition. San Diego: Academic Press. 413-468.

Machida, Sayuki. 2008. “A step forward to using translation to teach a foreign/second

language”. Electronic Journal of Foreign Language Teaching: 5, 1, 140–155.

Mackey, Alison, Al-Khalil, Maymona, Atanassova, Gergana,  Hama, Mika, Logan-Terry,

Aubrey & Nakatsukasa, Kimi. 2007. “Teachers’ intentions and learners’ perceptions about corrective feedback in the L2 classroom.” Innovation in Language Learning and Teaching: 1, 1, 129-152.

Malmkjær, Kirsten (ed.). 1998. Translation and Language Teaching: Language Teaching

and Translation. Manchester: St Jerome.

Schmidt, Richard. 1990. “The role of consciousness in second language learning”.

Applied Linguistics: 11, 129-158.

Schmidt, Richard. 1995. “Consciousness and foreign language learning: A tutorial on the

role of attention and awareness in learning”. Richard Schmidt ed.  Attention and awareness in foreign language learning. Honolulu: University of Hawaii Press. 1-63.

Schmidt, Richard. 2001. “Attention”. P. Robinson ed. Cognition and second

            language acquisition. Cambridge: Cambridge University Press. 3-32.

Takimoto, Masato & Hashimoto, Hiroko. 2010. “An ‘Eye-Opening’ Learning

            Experience: Language Learning through Interpreting and Translation”. Electronic

            Journal of Foreign Language Teaching: 7, 1, 86–95.

ZhaoHong, Han. 2001. “Integrating corrective feedback into communicative language

            teaching”. Academic Exchange Quarterly: 1, 23-32.