Published in the Canadian Modern Language Review - Volume 55, No. 4, June / juin 1999

To see more articles and book reviews from this and other journals visit UTPJOURNALS online at UTPJOURNALS.com

What's Wrong with Oral Grammar Correction

John Truscott

Abstract: The practice of oral grammar correction continues to receive wide acceptance in language classrooms. In this paper I argue that this acceptance is not justified. Teachers and students who take correction seriously face overwhelming problems, both in making the corrections effective and in dealing with the harmful side effects of the practice. Research evidence points to the conclusion that oral correction does not improve learners' ability to speak grammatically. No good reasons have been advanced for maintaining the practice. For these reasons, language teachers should seriously consider the option of abandoning oral grammar correction altogether.

Résumé : La pratique de la correction de la grammaire à l'oral continue à être bien acceptée dans la classe de langue. Dans cet article, cependant, je voudrais montrer qu'une telle pratique n'est pas justifiée. Les professeurs.es et les étudiants.es qui prennent au sérieux la correction d'erreurs rencontrent de graves problèmes à rendre la correction efficace et à réparer les effets néfastes de cette pratique. Les résultats des recherches antérieures se rejoignent à la conclusion que la correction à l'oral ne peut pas améliorer la capacité des apprenants de parler correctement. Les preuves avancées pour soutenir la pratique de la correction dans la classe n'y réussissent pas. Les enseignants de langue devraient donc considérer sérieusement la possibilité d'abandonner totalement la correction de la grammaire à l'oral dans l'enseignement de langue.

Introduction

Grammar correction remains a popular teaching practice in both written and spoken contexts. Its use has declined in recent years as a result of increased concern with communication, but the basic assumption underlying the practice is still largely taken for granted - that correction plays an important role in the development of students' ability to speak and write accurately. For most teachers, the issue continues to be one of details - when to correct errors and how to correct them. Given this continuing focus on details, one can easily forget that there is a more fundamental question: Should it be done at all?
For written correction, the case for a negative answer was recently made in Truscott (1996). I wish to take a similar look at correction in oral contexts, concluding that it is no more appropriate than written correction. I will first consider the many problems faced by teachers in their efforts to provide helpful correction and by students in their efforts to learn from it; I will then examine the research evidence regarding the effectiveness of the practice and the question of why oral grammar correction continues to be practised, despite its many problems. This is followed by a look at the question of whether correction might someday become a viable practice.
Throughout, the discussion will be restricted to errors in grammar. A similar case could be made for other types of errors (e.g., in pragmatics or pronunciation), but I will not attempt to do so here. It should also be emphasized that my thesis does not extend to feedback that falls under the heading of negotiation of meaning or negotiation of content, in which the goal is to facilitate conversation. It does include most of what has been called negotiation of form, which serves a didactic purpose (Lyster & Ranta, 1997; Van den Branden, 1997).

Teachers' problems

Teachers wishing to give helpful correction face a number of serious problems. Some are problems that get in the way of effective correction; others are undesirable side effects of correction, or of attempts to deal with the first type of problem.

Understanding the error

In order to provide effective correction for a student's error, the teacher must first determine exactly what that error is. Understanding is often made difficult by the inherent complexity of grammar, which can be challenging even to experts. Most language teachers are not experts on grammar, so problems are inevitable. For teachers who are not native speakers of the language, the problems can be compounded. And even expert knowledge cannot guarantee that the teacher will understand the most important aspect of the error - why the learner committed it. A correction based on a misunderstanding of the error's source could do more to confuse than to enlighten the student.
The process is further complicated by the context in which the teacher works. Even under the most favourable circumstances, dealing with grammar issues can be difficult; teachers commonly work in circumstances that are far from ideal for careful analysis of errors. Spoken errors occur quickly, often amidst extraneous noise. Learners' pronunciation is often unclear. Their limited ability to express themselves in the target language sometimes creates uncertainty about the intended meaning, making it difficult for teachers to determine exactly what error(s) occurred.
Additional problems come from the many competing demands on teachers' attention. While dealing with grammatical errors, they must also pay attention to the content of the utterance they are correcting and be prepared to respond to it. They must monitor other students' understanding and attentiveness, as well as their reactions to what has been said. They must also keep in mind the goals of the current activity and make any adjustments that become necessary.
A teacher who wishes to correct grammar errors under these circumstances also runs the risk of correcting a non-error. Having to deal with unclear speech, and distracted by other concerns, the teacher may at times hear a grammatical utterance as ungrammatical. The false correction that results will create further problems for students.

Presenting the correction

When teachers fully understand an error, they are still faced with the problem of clearly presenting the correction, along with any necessary explanation. Again, this must be done in the context of ongoing activities, so other factors will also make their claim on the teacher's attention. And, once again, the correction will often have to be designed to fit the reason for its occurrence. For a given error there may be a variety of possible causes and a variety of possible repairs; choosing one that is appropriate is not an easy task.
In presenting a correction, teachers must also be concerned with what the student can and cannot understand. Students have limited knowledge of grammar, so even when the teacher fully understands an error and presents what would seem to be a clear correction, the correction may fail because the student does not understand it. Even when the student does understand, this understanding may not extend beyond the particular context in which the correction occurred. The complexity of grammar often makes such generalization difficult. Presenting corrections in a way that overcomes these problems is challenging, to say the least.
The teacher must also deal with the question of whom the correction is aimed at. One student made the error, but others will hear the correction. A correction designed for one particular student may be quite inappropriate for the others. On the other hand, a correction designed to benefit the group as a whole may not be appropriate for the student who actually made the mistake.

Being consistent

If teachers are inconsistent in their corrections, these corrections are as likely to be harmful as they are to be helpful. Inconsistency can mean correcting one instance of an error and ignoring other instances; it can also mean offering two different types of correction for a single type of error, or, more precisely, what a student perceives as two different types of correction for what the student perceives as a single type of error - identifying error types is far from trivial, and the teacher's analysis may well conflict with the student's. Thus, the complexity of grammar is an important source of consistency problems, especially because teachers must identify the errors in the context of ongoing class activities.
Another source of problems is the failure to notice errors. People's ability to note the presence of an error cannot be taken for granted, even in ideal circumstances. Consider, for instance, the number of times that readers fail to notice typos, even when looking for them. In the midst of ongoing class activities, it is especially problematic. Teachers must deal with fleeting utterances, often spoken unclearly and inappropriately, in the midst of extraneous noise. At the same time, they must attend to the content of the utterance, prepare a response, and monitor other students' reactions, without losing track of the overall goals of the activity. These circumstances will inevitably produce oversights, probably a great many of them. Thus, many cases occur in which teachers unknowingly fail to correct errors that they do correct in other cases.
The situation is further complicated by the fact that correction rarely affects just one student, in isolation from the rest of the class. A teacher might have good reason to correct a certain student in one way and another in a different way (or to correct one and not the other). But all students will hear the varying corrections (or lack of correction). Thus, a reasonable concern with the needs of individual students will result in additional inconsistency for all students.
A final problem is the amount of effort required by teachers for even a moderate level of consistency. Even if consistency is possible, the effort to achieve it will greatly exacerbate the problems of the teacher who is already strained to deal with the basic problems described above.

Tailoring the correction to the student

Probably the most difficult aspect of the correction process is tailoring corrections to individual students. Teachers who wish to give effective feedback must consider its effect on each individual student. The problem has two aspects: the affective and the cognitive.
On the affective side, learners clearly differ in their reactions to correction. For some, no adverse effect is likely unless the corrections are delivered in a very aggressive or unfair manner. For others there is a serious danger that correction will produce embarrassment, anger, inhibition, feelings of inferiority, and a generally negative attitude toward the class (and possibly toward the language itself). To make correction effective and avoid harmful side effects, the teacher must see each student as a unique puzzle, asking how that student will respond to correction in its many possible forms, varying, for instance, in the type of error corrected, the frequency of correction, the explicitness of the correction, the amount and type of accompanying explanation, and the forcefulness of the correction. Determining how each student in the class is affected by each type of correction is a difficult project, and one that is certainly subject to error.
Even if the teacher has accurately assessed each student's likely reactions and successfully adjusted to them (an unlikely prospect), another problem arises: a serious concern with individual differences rules out consistency in correction. A teacher who ignores an error made by one student and then corrects the same mistake when it is made by another is sending a mixed and confusing message to the class as a whole. The teacher may also be perceived as overly critical of some students and overly kind to others. Another danger is that students will get wrong ideas about their ability, or about the teacher's perception of their ability. In particular, those who are corrected more often than others may conclude that they are inferior students, or that the teacher believes they are.
Despite these difficulties, affective problems are far more tractable than those on the cognitive side. Considerable evidence has accumulated that much grammar acquisition occurs in a relatively fixed order; learners are not able to master one aspect until after they have mastered certain others (for reviews, see Ellis, 1994; Larsen-Freeman & Long, 1991). Attempts to teach, via correction, an aspect for which the student is not yet ready are therefore likely to be ineffective, and quite possibly harmful. To provide helpful correction, a teacher must carefully and repeatedly test students' levels on each of a large number of different constructions and adjust the corrections accordingly.
This extremely complex task is further complicated by limits on what is now known about the order in which structures are acquired. Even for the languages that have been studied most thoroughly, current knowledge is too limited to provide guidance on more than a fraction of the errors teachers encounter. Prospects for success are also limited by the difficulty, in the context of ongoing classroom activities, of keeping track of which students should be corrected on which points. Dealing with the problem for a single student is quite impractical; doing it for an entire class of students is for all practical purposes impossible (for some discussion, see Ellis, 1993).
Once again, the effort to deal with students as individuals complicates the effort to be consistent. Corrections that are appropriate for one student will not be appropriate for others

Maintaining a communicative focus

Most teachers now share the assumption that effective instruction requires a communicative focus in the classroom. But reconciling this assumption with the use of correction creates difficulties for the teacher.
Correction, by its nature, interrupts classroom activities, disturbing the ongoing communication process. It diverts the teacher's attention from the essential tasks involved in managing a communicative activity. It moves students' attention away from the task of communicating. It can discourage them from freely expressing themselves, or from using the kinds of forms that might lead to correction.
To avoid these and other problems, or at least reduce them, teachers may refrain from overt correction and instead recast students' ungrammatical utterances in a grammatical form. This approach has its virtues, but it also introduces another source of confusion, as recasts are often used for purposes other than correction. A recast can be an alternative way of expressing the student's idea, an expansion, a simplification for the benefit of other students, an indication that the student did not speak loudly enough or did not speak clearly enough, or an attempt to stress the speaker's point to other students. The same is true for repetitions. Teachers frequently use them as corrections, but also as expressions of understanding or approval (Chaudron, 1988). The subtle differences between positive and negative uses are often undetectable to learners. Teachers who do not explicitly label their corrections, therefore, risk being misunderstood. Similarly, if students expect the teacher to correct them, they may well mistake a communicative recast or repetition for a correction. They will then believe that the original utterance was ungrammatical when in fact it was perfectly acceptable.
A similar problem arises with teachers' indications of approval. In the communicatively oriented classroom, teachers often use expressions of approval to indicate understanding of something a student has said, or to say that it was a good point, or to show agreement, or simply to praise the student for making an effort at self-expression in the target language. Students who expect the teacher to correct their grammar errors may mistakenly take such approval as an indication that their utterances are fully grammatical. Such confusion is much less likely if the teacher never ties expressions of approval and disapproval to grammar points.

Summary

Oral grammar correction is an extremely complex process, as can be seen, for example, in Chaudron's (1977) model. It is not surprising, then, that teachers who wish to provide effective grammar correction face enormous problems. Some of the difficulties are at least partly manageable, at least for a skilled teacher who is very serious about dealing with them. Classroom studies indicate, however, that teachers are generally not very successful in their attempts to provide high-quality oral corrections (e.g., Chaudron, 1988; Fanselow, 1977).
To reduce the problems, teachers might use delayed correction, recording students' speech or taking notes during class and then presenting them to students outside of normal class activities. One drawback of this approach is that it removes the error from its context, thereby reducing the relevance of the correction. More importantly, it makes the process very similar to written correction, the ineffectiveness of which has been reasonably well established (Truscott, 1996).
I have assumed in this discussion that corrections are made by the teacher. If the popular alternative of peer correction is used, most of the problems are unaffected. Some are exacerbated: peer correction moves students' attention further away from communicative activities, as well as producing corrections of lower quality. Porter (1986) found that learners rarely corrected one another's mistakes (suggesting that they only made corrections they were certain of), but that when they did so they were wrong one-sixth of the time.
The problem of tailoring corrections to the cognitive needs of individual students is simply beyond the ability of even the best and most dedicated teacher. The other problems, even if managed effectively, place a tremendous burden on the teacher, if taken seriously. One might ask why a busy teacher should devote so much time and effort to mastering the use of a technique that has so little prospect of success. The time and effort would be better spent on other aspects of teaching.

Students' problems

The problems faced by students in dealing with teachers' corrections are no less serious than those faced by teachers in providing good correction. In discussing these problems, I will omit certain types that I have already discussed from the teacher's perspective, such as the danger that correction will inhibit students' subsequent efforts at communicating.

Noticing and recognizing the correction

The first problem for students is to notice the correction and to recognize it as such. This aspect cannot be taken for granted. Students involved in an oral activity have other things on their minds, including the ideas they want to express, ways of expressing them, the things other students have said, and the way those students are reacting now. They may not hear the teacher clearly, possibly because of other sound in or near the room. They may hear but not understand well enough to know they have been corrected. If the teacher does not explicitly label a correction as such, they may mistake it for something else, as described above. Accompanying this problem is that of students taking the teacher's comment as a correction when it was intended for another purpose.
Recognizing the correction is only one of several requirements for the learner. In order for a correction to be effective, each of these requirements must be met, along with all those related to teachers. In this context, even a fairly low percentage of recognition failures would be a serious problem. Classroom research suggests that the percentage is actually very high (Roberts, 1995).

Taking the correction seriously

The student who has identified a correction must then pay attention to it, trying to understand it. This task may have to compete with others for that attention. The student may want to say something else immediately, or to listen to another student's comment, or to think about the topic of the discussion or about other aspects of the language forms being used. Nervousness or embarrassment may get in the way. The corrected point may be considered too difficult or too unimportant to bother with. Or the student may simply be unwilling to make the necessary effort.
In order for students to benefit from correction, they must be very serious in dealing with the corrections they receive. Research on writing classes indicates that most students are not so serious (see Truscott, 1996, and references cited there), despite the fact that they think teachers should correct them (e.g., Schulz, 1996). The problems in treating corrections seriously are in fact much greater in the oral context. In the midst of ongoing class activities, students will be much less able to maintain a focus on the correction, to give it the thought that may well be required if they are to understand it and remember it. The presence of other students observing the correction and the student's reaction to it also creates anxiety, causing further distraction.

Understanding and accepting the correction

Making an effort to understand is not sufficient; success in understanding is also necessary. Once again the complexities of grammar, the limits of students' knowledge, assorted distractions, and the student's own feelings can all get in the way. And a shallow understanding is not sufficient. In order for a correction to be useful, the student must understand the point well enough to be able to deal with it in a variety of contexts, beyond the one in which the correction occurred. Efforts to understand are further complicated by the fact that the corrections are fleeting - there is no written record, unless students are serious enough to take the time and effort to make a record. For the rare cases in which this occurs, the problems of maintaining a communicative focus are exacerbated.
A student who understands a correction must also be willing to accept it. The teacher's authority will go some way toward securing this acceptance, but it may fall short if the point conflicts with the student's intuitions, or if the student remembers (perhaps mistakenly) previous teaching or experience that seems to contradict what the current teacher is now saying. Students may also feel some resistance to the experience of being corrected. People in general do not enjoy being told that they are wrong; for students who must constantly deal with public correction, the experience can be quite distressing.

Incorporating the correction

Even if all of the above problems are overcome, the correction may still have little or no value because of the final, most important, problem. If a correction is to have a long-term impact on the learner's use of language, the information it conveys must be incorporated in the developing interlanguage, making possible its accurate, automatic use in the future. The alternative is for the learner to consciously note and control each future use of that information, a strategy that might have some value for a very small number of relatively simple grammar points but cannot possibly serve as a general basis for competent speech. But large questions arise as to how - and whether - incorporation of corrections occurs.
A common feeling among advocates of correction is that for a given instance to be incorporated (i.e., to have a lasting effect) the student must not only notice and understand it, but also deliberately rehearse it and make use of it. Fulfilment of this requirement runs into all the problems discussed above. Another requirement frequently mentioned is learner readiness - the student must be at a stage that is suitable for the particular point that is being corrected. As noted above, this condition is usually not met.
Even when this and all the other conditions are satisfied, however, the outcome is highly uncertain. The incorporation process is so poorly understood that one cannot say with any confidence how - or whether - explicitly acquired knowledge is incorporated. One viable position is that it is not possible at all, that the development of interlanguage grammar is one thing and the knowledge acquired through methods such as correction is another and that the two do not meet. This position is quite defensible, both theoretically and empirically (see, for example, Krashen, 1982; Paradis, 1994; Schwartz, 1986; Truscott, 1998; Zobl, 1995).
But this strong view is not a necessary condition for doubts about the incorporation of corrections. One might also take a weaker view, that explicitly acquired knowledge can be incorporated but that the process is difficult and perhaps requires special circumstances, or can only occur with certain types of grammatical knowledge. The conclusion that correction is important to learning follows only from strong views of another sort, namely from theories claiming a relatively simple and routine incorporation of explicitly acquired knowledge.

Summary

In order for a given instance of correction to be effective, both teachers and students must overcome a great many problems, any one of which is sufficient to make the correction a failure. While it would be too much to infer that no instance of correction is ever successful, one can infer that the chances of success in any given instance are vanishingly small. The occasional (accidental) success cannot begin to justify the problems created by the practice or the efforts required of teachers and students who attempt to take correction seriously.

Evidence on the effects of oral correction

Not surprisingly, available evidence suggests that oral correction is ineffective. In evaluating the research on correction, one must keep in mind three important points. First, success does not mean that students acquire the ability to solve grammar problems; research (and common experience) shows that this ability has no clear relation to the ability to speak or write grammatically (e.g., Frantzen, 1995; Kadia, 1988; Schumann, 1978; Terrell, Baycroft, & Perrone, 1987). Second, success cannot be judged by the student's immediate response to the correction; one cannot infer from any initial response that a correction has been, or ever will be, incorporated in the mental grammar. Third, success cannot be measured simply by students' use of the corrected forms in contexts that require their use. A common research finding is that learners who have been taught to use a particular form tend to overuse it (Lightbown, 1983, 1985, 1987; Lightbown, Spada, & Wallace, 1980; Pica, 1983; Weinert, 1987). Thus, tests that look only at uses of the corrected forms in obligatory contexts can greatly overestimate learners' success. A fair assessment of the results of instruction/correction must measure not only correct uses but also inappropriate uses.
Studies that avoid these basic problems (as well as some that do not) have found oral correction ineffective. Ellis (1984) carried out a brief study on the acquisition of wh- questions, finding that amount of correction did not appear to have any relation to success. Plann (1977) reported a much more extensive effort by teachers to correct the grammar of Spanish speaking children in an English immersion program; the efforts failed. Similarly, VanPatten (1986, 1988) reported the findings of two extensive studies by Dvorak, primarily on oral correction, which found that the practice was not helpful. Lightbown's (1983) extensive study of the development of English inflectional affixes obtained similar results: very serious long-term efforts by teachers to correct their students were accompanied by minimal gains in the students' accuracy.
In Van den Branden's (1997) study of native- and non-native-speaking children in a Dutch class, one experimental group received negotiation of form mixed in with negotiation of meaning and content. Afterwards the students' accuracy declined considerably. The author saw the decline as simply a consequence of the fact that they did more talking after the negotiation and therefore took more risks. Whether or not this is an adequate explanation, the results do not fit well with a claim that form-oriented feedback contributes to accuracy. This study was short-term, but was conducted under very favourable conditions: all negotiation of form took place in one-on-one situations with the researcher, who was thus able to control the interactions and adjust feedback to the needs of individual students, as well as avoiding the affective problems that can accompany public correction.

Perhaps the most interesting study is DeKeyser's (1993), which also found no overall effect for correction. But DeKeyser was more interested in possible interactions between correction and various individual difference variables. For tests that measured actual ability to use the target language, he found a significant interaction for only one of these variables, extrinsic motivation. The expectation was that correction would be especially helpful for students with high extrinsic motivation (a strong concern with rewards and punishments). The opposite was found. DeKeyser argued that this surprising result was due to the accidental involvement of another variable, anxiety. If one accepts this reasonable explanation, the study actually found no interaction between correction and extrinsic motivation (or any other factor).
Taken together with the other results, this non-finding indicates that DeKeyser (1993) found no relation of any kind between correction and learners' ability to use the target language (though he did find a relation between correction and performance on a grammar test). Further evidence of correction's irrelevance comes from the inclusion of grammatical sensitivity as an independent variable. No interaction was found with correction; in other words, students who were especially sensitive to grammar gained no more from correction than did students who were especially insensitive. It is difficult to imagine a clearer indication that grammar correction is irrelevant to learning.
Research on the effects of written correction is also relevant. This research consistently finds it ineffective (Truscott, 1996). Given this failure, one would expect oral correction to fail as well, since it poses even greater problems both for teachers and for students.
Recently, Ellis (1998) rejected the conclusion that correction is ineffective, based on the findings of Lightbown and Spada (1990) and Doughty and Varela (1998).1 One problem with Ellis's argument is that it gives more weight to two studies apparently supporting correction than to the far more extensive research that has found it ineffective. Another is that it overlooks crucial limitations of the two studies. Lightbown and Spada relied on limited, post-hoc observations, used no control group, and did not look at errors of overuse. In their words, the results `can only be taken as suggestive of directions for future research' (442).
Doughty and Varela (1998), studying ESL science classes, concluded that correction led to significant improvements in learners' use of past tense forms, which were maintained two months after the feedback ended. One major problem is that the testing was done in a very specialized way, closely corresponding to the way in which most of the feedback was presented during the study. Students answered five or six questions about science experiments they had just completed, the same questions that had provided the material for correction throughout the experiment, in nearly identical contexts. Thus it is not clear whether their improved use of past tense forms reflected a general change in their competence or the development of a specialized ability to deal with these specific questions in this specific context. Part of the problem is that students presumably knew that their use of past tense forms was being monitored during the testing, so their attention was probably focused on those forms. This fact raises serious doubts about the test's relevance to their normal use of English.
An even greater problem is that the testing procedure and analysis ruled out any measure of overuse - the type of error most likely to be committed by the corrected students. The questions were designed to produce contexts in which past tense forms would be obligatory, minimizing the opportunities for learners to overuse the forms. During the analysis, the authors found that one of these questions was not successful enough at creating obligatory contexts, so they discarded it. Even with these procedures, some overuses apparently remained in the data. These were eliminated by means of a scoring system that classified them not as errors but as signs of progress, apparently on the assumption that these errors would later be replaced by correct uses, while those involving under-use of the forms would not. It would be conservative to say that the validity of this assumption has not been established.

Why is oral grammar correction still widely practised?

Given all the reasons for skepticism about the value of correction, one might ask why it remains such a common practice. Is there some good reason for continuing to correct grammar errors, in spite of the problems? I will suggest several reasons for the continuing popularity of correction, concluding that none provides any justification for this popularity.
One reason, no doubt, is tradition. Grammar correction has been so much a part of language teaching for so long that its presence is largely taken for granted; little or no need is seen to justify its use. When such a need is seen, it is satisfied by research that provides only superficial support; that is, research that looks only at learners' immediate reaction to a correction or that finds correction helpful for performance on grammar tests. Neither type provides any evidence that correction makes learners better able to use the language in practical ways.
The continuing tradition probably owes a great deal to the dead hand of Behaviourism. In its heyday, the idea that learning means forming proper stimulus-response bonds became entrenched in teaching practice, in the form of the audio-lingual approach. A crucial part of the idea was that learners must not be allowed to utter incorrect sentences, as this would lead to (or strengthen) wrong stimulus-response connections. Prompt correction was therefore considered vital. The theory has long since been cast aside, but some of the effects of its once-powerful influence no doubt linger on.
More contemporary theories may also be affecting teachers' attitudes toward correction, though the extent of their influence is difficult to judge. Most important is the idea of hypothesis testing. If learning grammar means (in part) forming and testing hypotheses about the target language, then one can certainly imagine correction playing an important role, by providing evidence that a given hypothesis is false. But this idea rests on the questionable assumption (see above) that incorporation of explicitly acquired information is relatively straightforward. Even if one accepts this view and the claim that correction can - in principle - play an important role in hypothesis testing, overwhelming practical problems remain, as discussed earlier. So the appeal to hypothesis testing does not provide a valid basis for the continuing use of grammar correction.
Another powerful factor in the continuing popularity of correction is the common intuition that correction should work, even must work. Students say something the wrong way; they are told how it should be said; they then say it the right way. But the above discussion indicates that this simple view bears little resemblance to what actually happens in the correction process. In this case, intuition is more misleading than helpful.
The intuition that correction is effective is intimately associated with teachers' and learners' perceptions of their own experience - most people who have studied or taught a second language can point to cases in which a particular correction seemed to help. Given standard intuitions and tradition, it is an easy step to the conclusion that correction is generally helpful. There are two problems with this inference.
First, a great many cases of apparent benefits are probably illusory. Competent use of grammar is almost entirely unconscious; learners' ability to judge the effects of correction on their own production is therefore limited. What they are most aware of (almost by definition) is their metalinguistic knowledge, which is easily affected by correction but has only a very limited relation to actual language use. Thus, it is easy for learners to believe that a correction has helped them when in fact it has had no effect on their performance, or only a short-term effect, for the period in which they consciously thought about that particular correction.
The second problem is that even if cases of beneficial effects do occur, it does not follow that correction has any general value. Nothing in the discussion here suggests that no instance of correction is ever helpful; there may well be isolated cases in which benefits occur. The point is that if such cases occur, they are rare - too rare to show up statistically in the research - and are, for all practical purposes, random. There is no reliable means of predicting when (if ever) a correction will be effective. So such cases, if they do indeed exist, cannot begin to justify the use of correction.
A final reason for the continued popularity of grammar correction is the concern with students' attitudes. Most students strongly believe in correction (e.g., Schulz, 1996), so many teachers are understandably concerned about the effect that its absence will have on students' attitudes, and therefore on their learning. But available evidence suggests that this concern is inappropriate. Research on correction does not find uncorrected students less successful in their learning than corrected students. Nor has any evidence emerged that uncorrected students have attitude problems. Given the above discussion, one could plausibly argue that correction itself harms students' attitudes and that its absence is therefore beneficial. My own teaching experience indicates that students are quite capable of adjusting to the absence of grammar correction, even those who are initially very strongly convinced of its value.

Future prospects for oral grammar correction?

There is another possible reason for the continuing use of correction, particularly by teachers who are aware of its problems. Because correction involves a great many variables, the possibility remains that some untested combinations of these variables could produce successful feedback, while avoiding (or minimizing) the accompanying problems.
The outcome of correction may well be affected by a variety of learner variables, for example, proficiency, age, gender, or educational background. Also relevant is the nature of the class; the degree of control the teacher exercises over class activities, for example, may affect the success of correction. A variable that has received considerable recent attention is feedback type. Lyster and Ranta (1997), for instance, distinguish six types of feedback - explicit correction, recasts, elicitation, metalinguistic feedback, clarification requests, and repetition - and argue that some are better than others.
Perhaps the most interesting variable is error type. On theoretical grounds, one might expect certain types to be more amenable to correction than others. If one assumes principles and parameters theory, a distinction can be drawn between core grammar, the acquisition of which is tightly controlled by Universal Grammar, and peripheral items, not so tightly controlled. The latter should be relatively amenable to explicit teaching, including correction (Schwartz, 1993); i.e., incorporation of a correction may be more likely for peripheral targets (e.g., verbal inflection) than for core targets (a fundamental principle of syntax).
On practical grounds, simple errors should be more correctable than more complex errors, as they are relatively easy for teachers and learners to recognize and understand. Correction of simple errors, requiring less time and effort, might also pose less of a problem for efforts to maintain a communicative focus in the classroom. Developmental sequences are also relevant. Certain aspects of grammar are not known to be involved in these sequences and should therefore make better targets for correction.
Given all these variables and their uncertain relationship to the effectiveness of correction, it is perhaps understandable that some informed teachers still consider correction appropriate. But the belief is nonetheless misguided. Everything now known about correction indicates that the practice is unhelpful and creates major problems. The idea that certain combinations of variables could produce beneficial correction remains hypothetical, as does the possibility that harmful side effects can be adequately managed.
There is little reason for optimism about the future prospects of correction. First of all, it is unlikely to ever play a central role in language learning. Theories of second language acquisition disagree on many things, but they share a common belief that learning is driven primarily by input. Thus teaching styles that emphasize correction must be seen - now and in the future - as relics of the behaviourist past. The question, then, is whether grammar correction might be able to play a significant supporting role at some point in the future.
Again, there is little basis for optimism. The limited research on learner variables has so far found no relation to the effectiveness of correction (DeKeyser, 1993). For types of feedback, Lyster and Ranta (1997) provide evidence that some types are more successful than others in getting a response from students, but the chasm remains between these findings and the claim that any type of feedback actually benefits the learning process. For the theoretical distinction between core and peripheral items, any particular formulation is open to challenge. Application to the classroom is also difficult, because language use involves a constant interplay of core and periphery; in the context of class activities, separating the two would be a great challenge for even the most knowledgeable and capable teacher. Classroom application of findings on developmental sequences is no less challenging (see above).
Thus, in general, correction should be considered a bad idea. If one can find good reason to think that it is beneficial in a particular case and that the benefits outweigh the accompanying problems, then correction is appropriate in that case. But to this point no one, to my knowledge, has shown that any such cases exist.

Conclusion

Oral correction poses overwhelming problems for teachers and for students; research evidence suggests that it is not effective; and no good reasons have been offered for continuing the practice. The natural conclusion is that oral grammar correction should be abandoned.
Certain aspects of this conclusion require clarification. First, I am not suggesting that grammar is unimportant. The point is that correction does not make any contribution to the development of grammatical speech. Second, the conclusion applies specifically to grammar. The issues involved in correction of errors in pragmatics or pronunciation, for example, differ in some respects from those I have considered here, so my conclusion should not be casually extended to those areas. Most importantly, this paper is not a rejection of teachers' efforts to negotiate meaning or content in interactions with learners. In the negotiation process, some incidental feedback on grammar is almost inevitable. For example, when a student describes an event using verb forms that create confusion about the time of the event, the teacher may well seek clarification. It would be perverse to suggest that teachers should avoid this sort of negotiation. The point is that the goal should be simply to achieve successful communication, not to make students' utterances grammatical, or to tell them how to do so.
Finally, the rejection of grammar correction requires some caution. As noted above, the possibility remains that future research will find certain correction techniques effective with certain types of students and certain types of errors in certain contexts. But until such benefits have been demonstrated and have been shown to outweigh the problems caused by correction, teachers should not feel that they must, or even should, correct their students' oral grammar errors. Correction-free instruction is not only a viable alternative; given what is now known about the subject, it is the preferred option.

John Truscott is Associate Professor in the Department of Foreign Languages and Literature at National Tsing Hua University, Hsinchu, Taiwan. His research interests include the effects of correction and form-focused instruction on second language acquisition, as well as the implications of research in linguistics and cognitive psychology for second language learning and teaching.

Notes

1 In fact, Ellis (1998) refers to an earlier version of Doughty and Varela's paper. I will discuss the later version, which was not available at the time his paper was published.

References

Chaudron, C. (1977). A descriptive model of discourse in the corrective treatment of learners' errors. Language Learning, 27, 29-46.
Chaudron, C. (1988). Second language classrooms: Research on teaching and learning. Cambridge: Cambridge University Press.
DeKeyser, R.M. (1993). The effect of error correction on L2 grammar knowledge and oral proficiency. Modern Language Journal, 77, 501-514.
Doughty, C., & Varela, E. (1998). Communicative focus on form. In C. Doughty & J. Williams (Eds.), Focus on form in classroom second language acquisition (pp. 114-138). Cambridge: Cambridge University Press.
Ellis, R. (1984). Can syntax be taught? A study of the effects of formal instruction on the acquisition of WH questions in children. Applied Linguistics, 5, 138-155.
Ellis, R. (1993). The structural syllabus and second language acquisition. TESOL Quarterly, 27, 91-113.
Ellis, R. (1994). The study of second language acquisition. Oxford: Oxford University Press.
Ellis, R. (1998). Teaching and research: Options in grammar teaching. TESOL Quarterly, 32, 39-60.
Fanselow, J.F. (1977). The treatment of error in oral work. Foreign Language Annals, 10, 583-593.
Frantzen, D. (1995). The effects of grammar supplementation on written accuracy in an intermediate Spanish content course. Modern Language Journal, 79, 329-344.
Kadia, K. (1988). The effect of formal instruction on monitored and spontaneous naturalistic interlanguage performance: A case study. TESOL Quarterly, 22, 509-515.
Krashen, S.D. (1982). Principles and practice in second language acquisition. New York: Pergamon.
Larsen-Freeman, D.E., & Long, M.H. (1991). An introduction to second language acquisition research. London: Longman.
Lightbown, P.M. (1983). Acquiring English L2 in Quebec classrooms. In S.W. Felix & H. Wode (Eds.), Language development at the crossroads (pp. 101-120). Tübingen: Gunter Narr.
Lightbown, P.M. (1985). Input and acquisition for second language learners in and out of classrooms. Applied Linguistics, 6, 263-273.
Lightbown, P.M. (1987). Classroom language as input to second language acquisition. In C.W. Pfaff (Ed.), First and second language acquisition processes (pp. 169-187). Cambridge: Newbury.
Lightbown, P.M., & Spada, N. (1990). Focus on form and corrective feedback in communicative language teaching: Effects on second language learning. Studies in Second Language Acquisition, 12, 429-448.
Lightbown, P.M., Spada, N., & Wallace, R. (1980). Some effects of instruction on child and adolescent ESL learners. In R.C. Scarcella & S.D. Krashen (Eds.), Research in second language acquisition: Selected papers of the Los Angeles Second Language Acquisition Research Forum (pp. 162-172). Rowley, MA: Newbury.
Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake: Negotiation of form in communicative classrooms. Studies in Second Language Acquisition, 19, 37-66.
Paradis, M. (1994). Neurolinguistic aspects of implicit and explicit memory: Implications for bilingualism and SLA. In N.C. Ellis (Ed.), Implicit and explicit learning of languages (pp. 393-419). London: Academic Press.
Pica, T. (1983). Adult acquisition of English as a second language under different conditions of exposure. Language Learning, 33, 465-497.
Plann, S. (1977). Acquiring a second language in an immersion classroom. In H.D. Brown, C.A. Yorio, & R.H. Crymes (Eds.), On TESOL '77: Teaching and learning English as a second language: Trends in research and practice (pp. 213-225). Washington, DC: TESOL.
Porter, P.A. (1986). How learners talk to each other: Input and interaction in task-centered discussions. In R.R. Day (Ed.), Talking to learn: Conversation in second language acquisition (pp. 200-222). Rowley, MA: Newbury.
Roberts, M.A. (1995). Awareness and the efficacy of error correction. In R. Schmidt (Ed.), Attention and awareness in foreign language learning (pp. 163-182). Honolulu, HI: University of Hawai`i, Second Language Teaching & Curriculum Center.
Schulz, R.A. (1996). Focus on form in the foreign language classroom: Students' and teachers' views on error correction and the role of grammar. Foreign Language Annals, 29, 343-364.
Schumann, J.H. (1978). The pidginization process: A model for second language acquisition. Rowley, MA: Newbury.
Schwartz, B.D. (1986). The epistemological status of second language acquisition. Second Language Research, 2, 120-159.
Schwartz, B.D. (1993). On explicit and negative data effecting and affecting competence and linguistic behavior. Studies in Second Language Acquisition, 15, 147-163.
Terrell, T.D., Baycroft, B., & Perrone, C. (1987). The subjunctive in Spanish interlanguage: Accuracy and comprehensibility. In B. VanPatten, T.R. Dvorak, & J.F. Lee (Eds.), Foreign language learning: A research perspective (pp. 119-132). Cambridge: Newbury.
Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language Learning, 46, 327-369.
Truscott, J. (1998). Noticing in second language acquisition: A critical review. Second Language Research, 14, 103-135.
Van den Branden, K. (1997). Effects of negotiation on language learners' output. Language Learning, 47, 589-636.
VanPatten, B. (1986). Second language acquisition research and the learning/teaching of Spanish: Some research findings and implications. Hispania, 69, 202-216.
VanPatten, B. (1988). How juries get hung: Problems with the evidence for a focus on form in teaching. Language Learning, 38, 243-260.
Weinert, R. (1987). Processes in classroom second language development: The acquisition of negation in German. In R. Ellis (Ed.), Second language acquisition in context (pp. 83-99). Englewood Cliffs, NJ: Prentice-Hall.
Zobl, H. (1995). Converging evidence for the `acquisition-learning' distinction. Applied Linguistics, 16, 35-56.

xxxxxxxxxxxxxxxxxxxx
© 1999 The Canadian Modern Language Review/La Revue canadienne des langues vivantes,
55, 4 (June/juin)