Spoken grammar awareness raising : Does it affect the listening ability of Iranian EFL learners ?

Advances in spoken corpora analysis have brought about new insights into language pedagogy and have led to an awareness of the characteristics of spoken language. Current findings have shown that grammar of spoken language is different from written language. However, most listening and speaking materials are concocted based on written grammar and lack core spoken language features. The aim of the present study was to explore the question whether awareness of spoken grammar features could affect learners’ comprehension of real-life conversations. To this end, 45 university students in two intact classes participated in a listening course employing corpus-based materials. The instruction of the spoken grammar features to the experimental group was done overtly through awareness raising tasks, whereas the control group, though exposed to the same materials, was not provided with such tasks for learning the features. The results of the independent samples t tests revealed that the learners in the experimental group comprehended everyday conversations much better than those in the control group. Additionally, the highly positive views of spoken grammar held by the learners, which was elicited by means of a retrospective questionnaire, were generally comparable to those reported in the literature.

Over the last 30 years, owing to the intensive work in different areas of applied linguistics, the view regarding superiority of literacy has been changed and the primacy of oral production has been realized.Consequently, greater attention has been paid to the problem of understanding the spoken form of the foreign language (Brown, 1990).Listeners, when listening to spoken texts, encounter a number of features unique to spoken discourse (Buck, 2001;Rost, 2002).They need to have enough knowledge of these features in order to comprehend and thus communicate effectively.To this end, McCarthy and Carter (1995) argue that learners should receive special instruction on the characteristics of spoken grammar and these features should be integrated into English as a foreign language (EFL) teaching materials.
In general, spoken grammar is viewed from three different perspectives.The first view rejects the exisistence of any grammar except written grammar (Leech, 1998), although nowadays, with advances in corpus linguistics and analysis of spoken corpora, this view is not taken seriously.The second view holds that there is no special grammar of spoken language, and its proponents (Biber, Johansson, Leech, Conrad, & Finegan, 1999;Cullen & Kuo, 2007;Leech, 2000) believe in the same grammar performing different functions for written and spoken language.They argue that speech and writing draw on the same underlying grammatical system rather than on two separate systems.However, some structures such as ellipsis are more common in speaking rather than in writing.The last view, however, maintains that there is a special grammar of spoken language.Carter and McCarthy (1995) highlight the grammatical features of spoken language largely neglected by standard grammars and somehow take the view of a new grammar of speech.Nevertheless, they state that "spoken language and written language are not sharply divided but exist on a continuum" (Carter & McCarthy, 2006, p. 164).As Wendy and Lam (2002, p. 250) point out, "spoken language is not written language spoken aloud," but it has its own features and structures that differ from written language.Carter and McCarthy (2006) define spoken language as an interactive, normally unplanned, face-toface process, full of pauses, repetitions, interruptions, ellipsis, discourse markers, vague language, and hedges.Similarly, Biber et al. (1999) describe structural and functional aspects of spoken language based on spoken corpora.
Interest in spoken grammar has arisen with the growing availability of large computerized corpora.The findings of the studies of spoken corpora have been reported in detail in many publications (Biber et al., 1999;Brazil, 1995;Carter & McCarthy, 1995, 1997, 2006).Although corpora are accepted as a valuable source of authentic language use for both research and pedagogy, there are still debates on their application in language pedagogy.
The key controversial issue is whether teaching materials should be corpus-driven or corpus-based.Some corpus linguists (Sinclair, 1991;Stubbs, 1996) emphasize using authentic corpus-driven materials in language pedagogy.On the other hand, Widdowson (2000) questions the authenticity of corpus-based materials claiming that authenticity is a function of text reception as well as text production.He argues that learners who are not able to create contexts similar to those which native speakers produce will have problems with processing the texts which are created for language learning as well.Another opponent, Cook (1998), believes that learners run the risk of "produc[ing] corpus-attested but contextually inappropriate language" (p.60).However, the proponents of corpus-based or corpus-informed materials (McCarthy & Carter, 1994;Summers & Rundell, 1995) accentuate that materials should be influenced and informed by corpus findings and the data should be modeled on authentic patterns; therefore, they should be corpus-based rather than corpus-bound.
Nonetheless, literature indicates that listening and speaking materials are actually based on the written language norms.Carter (1998) compared real data from the spoken corpus with textbook dialogues and realized that core spoken language features were absent from these dialogues.This is also supported by Cullen and Kuo (2007), who, after surveying 24 general EFL textbooks published in the year 2000, concluded that in these books "coverage of features of spoken grammar is at best patchy" (p.361).Mumford (2009) highlighted three approaches to spoken grammar: World Englishes/English as a Lingua Franca (ELF), the passive knowledge approach, and the production approach.The supporters of World Englishes/ELF (Prodromou, 1996;Rajagopalan, 2004;Widdowson, 1994) claim that there is no need for specifically native speaker norms.For example, referring to estimates that up to 80% of communication in English takes place between nonnative speakers, Prodromou (1996) concludes that corpus language is inappropriate to the needs and interests of learners.In the same vein, Rajagopalan (2004) asserts that simply mutual intelligibility between speakers is a desirable goal in communication.In contrast, Kuo (2006) maintains that learners' need is beyond merely international intelligibility; they should be allowed to follow native speakers' model in order to communicate effectively in native and nonnative contexts.In her research, Kuo (2007) comes to the conclusion that for many of her students, native speakers' model is desirable.

Approaches to Spoken Grammar
The passive knowledge approach, on the other hand, contends that learners should have a passive knowledge of spoken grammar, and in order to raise students' awareness of these features, recordings of native speakers' scripts and noticing tasks should be used.An advocate of this approach, Timmis (2005), provides a framework for spoken grammar teaching; however, he does not recommend its teaching for production.He argues that "it is quite hard to frame useful and digestible production rules" (p.120).Alternatively, the production approach claims that learners should be able to use native speakers' norms in their production; thus, there is a need to go beyond passive knowledge since learners who are unaware of these norms will suffer a distinct disadvantage when encountering native speakers.
An important question regarding the application of spoken grammar in the classroom is the way it should be practiced.McCarthy and Carter (1995) propose a "three I's" paradigm including illustration, interaction, and induction.According to this paradigm, first particular forms in real data are illustrated, then learners do some tasks which actively involve them in noticing features through interaction, and finally, learners induce the patterns of usage.Another model is proposed by Timmis (2005), who maintains that noticing tasks and activities can help learners produce the features.Although spoken grammar continues to be an unresolved issue, rules of speaking seem an inextricable component of communicative competence (Cameron as cited in O' Keeffe, 2009).Consequently, learners' awareness of the features can play a vital role in enabling them to communicate.

Spoken Grammar and Listening
Listening is the medium through which spoken grammar can be comprehended.Carter and McCarthy (2006) claim that spoken grammar should be heard and not just read from a written text.O'Keeffe, Carter, and McCarthy (2007) state that "listening to spoken grammar, along with noticing tasks, can raise learners' awareness of its features and this kind of listening is often best carried out as 'listening for something' rather than 'listening to something'" (p.137).They suggest that more advanced learners can be encouraged to notice and spot spoken language features in listening and then to discuss why they are different from the written norm.In turn, this will help them develop real-life listening skills.Buck (2001) defines L2 listening ability as the ability to process extended samples of realistic spoken language in real time, to understand the linguistic information included in the text, and to make whatever inferences are unambiguously implicated by the content of the passage.The objectives of listening instruction, according to Rost (2006), should focus on any of the four areas in-cluding "improving learners' comprehension of spoken language, increasing the quality of learners' intake from spoken input, developing learners' strategies for better understanding of spoken discourse, or engendering a more active participation in face-to-face communication" (p.47).In order to help learners develop effective listening skills, the methods of instruction and the type of materials are very important.Regarding the type of materials, as Johns (1994) points out, learners should be exposed to corpus-based materials to develop the appropriate use of a language.Learners who are exposed to such materials and learn some features of spoken grammar will be more familiar with characteristics of spoken discourse as well as the strategies speakers use to orient, monitor, manage, modify, and soften their message, the result of which will be better comprehension and more appropriate response to what is being said.
In the present study, the researchers have adopted the passive knowledge approach for teaching conversational spoken grammar and have employed awareness raising activities for improving conversational listening ability of Iranian EFL learners.This study contributes to the research on spoken grammar as a very novel field in ELT as well as to listening instruction studies.The significance of the study lies in the fact that research in spoken grammar is limited and much of the discussion in the field regards the role of spoken grammar in speaking ability of L2 learners rather than in the skill of listening.
The present study, accordingly, addressed the following research questions: 1. Can awareness raising of spoken grammar through corpus-based instruction enhance the listening comprehension ability of Iranian EFL learners?2. What is learners' attitude towards corpus-based materials and spoken grammar instruction?

Participants
The participants of the study were 45 male and female Iranian university students aged between 18 and 29 who were majoring in English translation at one of the major universities in Tehran, Iran.They were selected based on convenience sampling and were members of two intact classes randomly assigned to the experimental (n = 22) and the control group (n = 23).

Instrumentation
Three different instruments were utilized for gathering the data required for this study.The first instrument was the Preliminary English Test (PET), which was used to ensure the homogeneity of the groups and was piloted with 20 students whose proficiency level was identical to that of the participants of the study.The overall reliability of the test computed through KR-21 was 0.85 and the interrater reliability of the writing section of the test was 0.90, showing a high correlation between the scorings of the two raters.
The second instrument was a 30-item achievement listening test adapted from the self-listening comprehension parts of the corpus-based Touchstone (levels 3 and 4) by McCarthy, McCarten, and Sandiford (2006a, b) used as the pre-and posttests (Appendix A).The test was administered to the participants of the study before and after the treatment to determine whether there was any gain in the listening scores of the participants after the intervention.In the process of these two administrations, the "B-index" of the achievement test was computed "to make decisions about which items to keep and which to discard" in the criterion-referenced test (Brown, 2005, p. 84).The items which appeared to be acceptable remained in the revised version.The agreement of the test was computed by estimating the threshold loss agreement through the Subkoviak approach, which is usually used to estimate the reliability of criterion-referenced tests (Brown, 2005).The estimated agreement coefficient of the test was .84,which justified its use in the study.Also, its content validity was approved by two university instructors based on a table of specifications prepared by the researchers.
Additionally, a 30-item 5-point Likert type questionnaire in the native tongue of the participants was designed to probe students' opinions regarding the course.Its content validity was approved by two experts and its reliability estimated through Cronbach's alpha (r = .81)signified a relatively high reliability index (see Appendix B for the English translation of the questionnaire).

Materials
As mentioned earlier, the materials were chosen from corpus-based Touchstone (McCarthy et al., 2006a, b), which according to Ruhlemann (2008) is unique among English textbooks and employs the three I's methodology for raising students' consciousness of spoken grammar.It is worth mentioning that the term spoken grammar is synonymous with conversational grammar and refers to the grammar of informal, conversational English rather than to the discourse used in more formal settings such as debates or speeches.Moreover, it is different from vernacular or nonstandard forms of grammar (Biber et al., 1999), which are restricted to regional dialects and are sometimes regarded as a sign of ill-education (see Appendix C for features of spoken grammar).

Procedure
The participants attended a 16-week listening and speaking course.The classes met two times a week, each session lasting 90 min.Seventy minutes of each session were allocated to this study.The two classes had the same listening texts taken from Touchstone, but their tasks and activities were different.Both classes were taught by one of the researchers, which may be regarded as a limitation of the study.
In the first week the PET was administered to the participants and it was verified that the two groups were homogeneous at the onset of the study.Subsequently, the two classes were randomly assigned to the experimental and control groups.In order to guarantee that the participants were not familiar with the spoken grammar features, the participants of the two groups sat a listening pretest as well.
In Weeks 2-15 the participants in both groups practiced listening using the same materials in pre-, while-, and post-listening stages for 70 min per session.The control group did not receive any explicit instruction on the spoken grammar features and was implicitly exposed to them.To stimulate and generate background knowledge before the listening tasks, the group received prelistening tasks based on content related to the topic.Afterwards, as the whilelistening activity, the group listened to a dialogue and answered some listening comprehension questions or did fill-in-the-blank tasks.In the post-listening phase, however, the learners were engaged in different activities such as role playing, paired dialogue writing, and discussion concerning the topics.
In contrast to the control group, the three I's methodology, as awareness raising tasks for spoken grammar instruction, was used in the experimental group.To activate the learners' background knowledge, during the pre-listening stage some questions were put forward by the teacher to introduce the topic of the lesson.Afterwards, the learners listened to a corpusbased real-life dialogue adopted from Touchstone, the targeted features were illustrated through a noticing task, and the learners were asked to do some activities which enhanced their awareness regarding the target features (Appendix D).Finally, in the post-listening phase, the participants were encouraged to discuss the context-related functions of the feature.
In Week 16 the same listening test used as the pretest was administered to both groups as the posttest as the final session of the study.

Proficiency Test
As the first step, two classes were selected based on convenience sampling and took the PET, the results of which signified the homogeneity of the two groups in terms of their language proficiency level.This was done to keep the variable of language proficiency constant.Hence, the two classes were randomly assigned to the experimental and control groups.
Next, an independent samples t test was run to compare the mean scores of the groups on the proficiency test.The distribution of the scores on the PET was normal with respect to the skewness ratios, 0.256 for the experimental group and 0.846 for the control group, falling within the normality range of -1.96 and +1.96.It should be mentioned that the values are obtained from dividing the statistics by the standard error of skewness.The Levene's test p = .54verified the equality of the variances and thus, the legitimacy of running a t test.The results of the independent samples t test (t = -0.057,df = 43, p = .955)revealed no significant difference between the means of the two groups at the .05level of significance.

Pretest
To check the homogeneity of the two groups in terms of their conversational listening ability, a listening comprehension pretest was administrated to both groups, the results of which are offered in Table 1.As Table 2 illustrates, the result of Levene's test, p = .547,signified the equality of the variances and the t observed (t = 0.114, df = 43, p = .910)showed no significant difference between the means of the two groups on the listening pre-test indicating the same level of conversational listening ability before the intervention.

Posttest
After the intervention, the test used as the conversational listening pretest was administered to the participants as the posttest, the descriptive statistics of which are given in Table3.The results of the posttest (t = 2.036, df = 43, p = .048)revealed that there was a slightly significant difference between the conversational listening ability of the experimental and control groups.Moreover, as the table shows, the eta squared value ( 2) for variability in the listening ability (dependent variable) was .087,which means that 8.7% of the variability in the listening ability can be accounted for by the spoken grammar suggesting a moderate effect size (utilizing the commonly used guidelines proposed by Cohen, 1988: .01= small effect, .06= moderate effect, .14= large effect).

Questionnaire
As mentioned in the instrumentation section above, to find out about the opinion of the experimental group on the course, the members were asked to fill in a questionnaire with 30 items on a 5-point Likert type scale and also to write their comments about the course while remaining anonymous.The responses were scored as: strongly agree (5 points), agree (4 points), unsure (3 points), disagree (2 points), and strongly disagree (1 point).The proportion of students agreeing with each category in the questionnaire is presented in Table 5.
As the results show, a vast majority of students agreed that spoken grammar instruction was helpful, particularly as an aid for improving their listening ability (84%) and their familiarity with spoken language features (92%).Particularly noteworthy was the fact that a large majority believed that they needed to know the spoken grammar features and the way native speakers talk to each other (93%).(Qs 7,12,14,15,17,23) 7 Regarding the course efficacy, there was 86% agreement that the objectives of the course were achieved and the course was highly effective.In addition, about 83% of the students agreed that the tasks were useful and helped them to understand the materials better.Furthermore, the materials were regarded by students as highly in line with the course objectives and useful (88%).It was also believed by 85% of students that the course increased their motivation, self-confidence, and autonomy.

Discussion
As regards the first research question concerning the ability of awareness raising of spoken grammar through corpus-based instruction to enhance listening comprehension ability of EFL learners, the results from the independent samples t tests suggest that learners informed of the features in the experimental group comprehended the everyday conversations significantly better than the learners in the control group, although the p value was only slightly lower than the cut-off point at the .05level of significance for rejecting the null hypothesis.This shows that teaching spoken grammar features, in general, could be beneficial for understanding the everyday-life conversations of native speakers and employing noticing and awareness raising tasks could have some role in bringing about the improvement of the learners.However, the finding is in line with the theoretical basis of the study which claims that listening to spoken grammar along with noticing tasks helps learners become aware of what they are likely to hear and also know the intention of the speaker.It can be assumed that once learners can make predictions based on the functions of words and phrases and their interpersonal meaning, they become better listeners and communicate more effectively.The finding supports Carter and McCarthy (2006), who believe that spoken grammar should be taught through listening, and listening can be enhanced through raising learners' awareness of the spoken grammar features.
The students' positive views on teaching the spoken grammar features elicited by the questionnaire are generally comparable to those observed by Timmis (2002), Goh (2009), andKuo (2006) in their surveys.As they show in their studies, teachers and students are quite positive toward the need for teaching and learning spoken language features and confirm the necessity for paying special attention to the rules and strategies used by native speakers in their conversations.The usefulness of the noticing tasks similarly supports Timmis' ( 2005) claim that such tasks encourage learners to compare what they say with what a native speaker says in the real world and the finding that 92% of students his survey found the tasks useful or very useful.It is evident from the results that awareness of spoken grammar features motivates and, as Goh (2009) states, empowers students because they gain more selfconfidence to encounter native speakers.
It is worth mentioning that nine out of 21 of the students commented on the open-ended questions of the questionnaire.Their comments can be categorized as follows: 1. Pace of the course: Three students complained about the pace with which the courses were taught.For example, one wrote that "although all tasks were fully covered, the pace was too fast, I think there must be more sessions."2. Speaking: Another topic of comments was the speaking ability of the participants, which was said to be influenced by the course material and spoken grammar instruction.The following are examples of the comments: "I think I like to use the things I learnt when I am speaking; I unconsciously use something I learnt when I speak and this makes me happy; I used to be very careful in using correct grammar when I was speaking, but now I take it easy."3. Audiovisual materials: Two students suggested that they preferred to have video texts instead of listening texts.They wrote: "When I watch something I can better understand what is going on; It was better to use video in the class."The first comment points to one of the limitations of the study.Actually, more time was needed for gaining the optimum results.The second issue, speaking, is strongly related to spoken grammar and the comments denote its positive impact on the learners' speaking ability.However, the researchers delimited the study to the listening ability because the cultural and contextbased features inherent in spoken grammar are still considered to be controversial issues.Employing audiovisual materials, in fact, provides learners with the visual cues which can trigger appropriate schemata for understanding and can be an interesting topic for further research.
The overall results, although they cannot be fully generalized, imply that spoken grammar instruction may not only enhance the listener's abilities mentioned by Buck (2001), but may also cover the first three goals of listening instruction proposed by Rost (2006), discussed earlier in this paper.When learners are exposed to corpus-based materials modeled on corpus-driven conversational texts, and also when they have the opportunity to be aware of the features a native speaker employs in speaking, their comprehension of everyday conversations is facilitated.This reduces the burden of processing the flow of information they receive.These elaborated listening texts can enhance comprehension in the same way as simplified texts do without damaging the richness of the original text (Long, 1996).

Conclusion
As the findings of the present study divulge, since corpus-based spoken grammar instruction through corpus-based materials can have an impact on EFL learner's everyday listening skills and can enhance their ability in communication, its implementation in listening and speaking courses seems desirable.It seems that this type of instruction works well if corpus-based materials and consciousness raising activities are employed.However, it is in fact rarely used in EFL classrooms and debate about its pedagogic relevance is still continuing.More investigations are needed to fill the gap between the findings of this study and their application to language courses.Hence, the findings of similar studies can have important implications for the domain of language teaching and can aid those researchers who are interested in investigating how awareness of spoken grammar facilitates real-life listening skills.

Questionnaire
The following questionnaire is about the spoken grammar you have been taught, and your listening course.Remember there are no right or wrong answers.Just answer as accurately as possible.Show your agreement or disagreement with each sentence by marking one letter.A = strongly agree B = agree C = unsure D = disagree E= strongly disagree thinks that the battery . . .a. is replaced by the wrong one.c. is not fixed in its place.b. has no problem.d. has no power.30 -The clock is fixed by . . .a. K a y l a .c. Hector.b.Hector and Kayla.d. none of them.

Table 1
Descriptive statistics for the conversational listening pretest

Table 2
Independent samples t test for the conversational listening pretest

Table 3
Descriptive statistics for the conversational listening posttest

Table 4
Independent samples t test for the conversational listening posttest

Table 5
The proportion of students agreeing with each major category