Developing intonation skills in English: A systemic functional linguistics perspective

This paper explores whether the teaching of English intonation within the framework of systemic functional linguistics (SFL) contributes to the development of intonation skills of Argentine Spanish speakers to become teachers of English as a Foreign Language. The findings of the s tudy that focused on the oral produ ction of s tudents in the fi rs t course of phonetics in the programme offered at the National Universi ty of Rio Cuarto are presented. This paper reports the analysis of recordings of fi rs t-year s tudents reading an English text aloud and the results obtained in the preand post-tests reveal that there was improvement in s tudents ’ oral production such as in tone system considering the three systems of intonation in SFL after a series of training sessions. Hence, this approach seems promising for the developmen t of intonation skills and oral skills in foreign language learners . These results favour for teacher trainers as well as for trainers


Introduction
It has long been acknowledged that the teaching of intonation has not received the attention it deserves considering the central role it plays in communication, be it in non-native-native or nonnative-non-native communication. This lack of interest in the field of phonetics and phonology has been noticed by a number of researchers in different contexts (Clennell, 1997;Derwing & Munro, 2005;Roncero, 2009;Valenzuela Farias, 2013;Wennerstrom, 2001), so much so that the field has even been labelled 'the orphan of second language rese arch and teaching' (Derwing, 2010, p. 24). Diverse reasons have been proposed for this lack of the attention, particularly in relation to prosody. Clennell (1997) points out that the discourse/pragmatic functions of English prosody appear to be specific to the English language and not readily appreciated even by native speakers, and that there is the problem of first language interference and the lack of suitable materials for teaching English prosody. Lack of interest may also be attributed to the fact that the intonation system is difficult to acquire by non-native speakers of English (Ramirez Verdugo, 2002), and that teacher training programmes have neglected the teaching of pronunciation, which has resulted in a whole generation of teachers who grew up with inadequate skills to teach it (Levis, 2006). In our context, where English is taught as a foreign language, all teacher training programmes have at least three compulsory phonetics courses in their programmes; however, it has been noticed that the know ledge future teachers acquire about phonetics and phonology barely transfers to their teaching practice. Indeed, this is perceived as a problem not only by teacher trainers but also by students.
The marginality of research and teaching of intonation in second and foreign language classrooms have most likely led to the disregard of pronunciation and intonation in classrooms in different educational contexts. This, in our view, is a serious drawback not only for the field, but most importantly, for each and every foreign language student who is not given the full repertoire of tools that lead to successful communication.
In the literature reviewed, it has been observed that research on intonation has grown in recent years. Derwing and Rossiter (2003) studied the productions of ESOL students in three different instructional settings, one with emphasis on segments, another on prosody and the third one with no specific emphasis. Their results suggest that the instruction of pronunciation with a focus on prosody led to higher ratings in comprehensibility and fluency than instruction focusing on segmental features. These authors suggest that an overemphasis on segmental pronunciation can distract a learner's attention and actually lead to a loss of fluency in speech (pp. 12-13). Ramirez Verdugo (2005) reports a cross-linguistic computerised corpus study that focuses on the comparison of the prosodic forms produced by Spanish learners and English native speakers in order to interpret the pragmatic meaning expressed in the conversations of both language user groups. Her results reveal that the Spanish speakers' choice of the English tone system may lead to pragmatic incompatibility in the expression of modality in their interactions. Pickering (2009) studied the prosodic features of 25 proficient and nonproficient non-native speakers of English from a variety of L1 backgrounds in a university setting. She used the discourse-pragmatic approach to intonation proposed by Brazil (1997) to study students' productions in different interactive activities such as giving directions, information gap, dialogue reading and informal conversation. Her results show that participants seem to use pitch cues as signals of possible trouble source, for example, in the case of misunderstandings or when an interlocutor approaches an unknown lexical item. Also, the author reports that pitch cues are used as a means to indicate that negotiation or repair sequences have been accomplished successfully.
Similarly, a growth in research has taken place in the context of Argentina, where Spanish native speakers learn English as a foreign language (EFL). Despite the scholarly acknowledgement that there are specific aspects of the field of phonetics and phonology that have been neglected (Roncero, 2009), different researchers have focused on the teaching of intonation to Argentine students at university level by presenting proposals for the teaching of a specific aspect (Barbeito, Cardinali & Di Nardo, 2013;Germani & Rivas, 2011;Macagno & Piccirilli, 2009;Perez & Acosta, 2009), or alternative proposals for the assessment of students' production in a phonetics and phonology course for first year students at university level (Cosentino, 2011).
In our specific context, where most teachers of English are non-native speakers, it seems important to focus on research and instruction of intonation. Considering the fact that the teachers and language learners are native speakers of Spanish, it is important to consider specific training regarding intonation, since the L1 and L2 are characterised by different pitch contours and this can negatively affect the way that Spanish speakers of English perceive and apply the rules of English intonation (Valenzuela Farias, 2013). Some researchers have studied students' performance regarding one aspect of phonetics and phonology as in the study reported in Perticone (2009). This researcher focuses on the identification and classification of errors of English nuclear tones produced by five River Plate Spanish-speaking hearers when taking tonetic dictations. The results of her work show that the participants tend to misidentify certain tones (fall-rise, rise and rise-fall), and she presents a typology of misidentification of nuclear tones. Some other studies have focused on analys ing students' performance after being taught phonetics and phonology within the systemic functional linguistics (SFL) framework. In a study carried out in 2012, the results show that students' ability to identify tonality, tonicity and tone improved after treatment (Barbeito & Cardinali, 2013;Barbeito et al., 2013;Barbeito, Cardinali & Di Nardo, 2014).
Even though the literature reviewed shows that interest in research in the field is growing worldwide (Pickering, 2009;Ramirez Verdugo, 2005) and different studies in the field have been carried out in Argentina, there is absence of studies that focus on the teaching of intonation from a SFL perspective. Specifically, it has been observed that there are no studies on the effects that the teaching of intonation from an SFL perspective have on the development of oral production of university students who are native Argentine Spanish speakers in teacher training colleges. Taking this into consideration, this paper aims to answer the following research questions: i.Does the teaching of intonation from the perspective of SFL affect the performance of students during reading aloud in an Argentine teacher training university course on phonetics?
ii. If so, which system is most affected?
On the teaching of intonation, different authors have highlighted the value of the suggestions proposed by Allen in 1971, which include directing students' attention to major selected patterns, showing the differences between the punctuation systems and the intonation system, the differences in the intonation of isolated sentences and segments in extended discourse and the importance of considering the speaker's intention (Yamato, 2004). This list was extended by Celce -Murcia, Brinton and Goodwin (1996) who suggested that the notions of given and new information and similarities and differences between the L1 and L2 are important additional points. Considering that intonation is usually not self-evident to non-native speakers and that Spanish in particular uses a narrower pitch range than English (Celce-Murcia et al., 1996), the teaching of intonation is an essential component of any course of EFL. Foreign language speakers need to be aware of the fact that intonation is not marginal in oral communication and that it serves different purposes: it organises information, it realises different communicative functions, it expresses attitude, it serves to identify syntactic structures, it contributes to signalling the textual structure, it characterises different speech styles or genres and it serves a sociolinguistic function (Tench, 1996).
Taking this into account, in this paper, it is argued that the specific teaching of intonation within the framework of SFL is useful for future non-native EFL teachers, and that work in this line may hopefully result in their renovated interest in the field both for research and teaching practice.

Theoretical framework
From the perspective of SFL (Halliday, 1994;Halliday & Matthiessen, 2014), language is a complex semiotic tri-strata system that comprises different levels of realisation: the semantic level (meaning), the lexico-grammatical level (words, groups or clauses) and the phonological level (sounds), which work as sets of systems that offer language users a variety of choices by means of which they construe their experience, interact and establish relationships with others, and organise the information flow (Halliday, 1994). In this framework, speakers make their language choices based on the contexts of culture and situation. The context of culture concerns the socio-cultural context where language is used and serves the purposes of communication, the ideology, conventions, institutions and social purposes. It is at this level that the concept of genre is defined as a social process that is goal-oriented and realised in different stages (Martin & Rose, 2007). The second context, the context of situation, concerns the situations that are framed in the broader context of culture and are realised in the register through its variables: field, tenor and mode. These variables, which are realised in lexicogrammatical configurations that characterise a particular text as a communicative event, are the ones that simultaneously map the metafunctions of language: ideational, interpersonal and textual.
SFL considers the phonological stratum as the level of expression of language where lexicogrammatical choices realise the stratum of content (Halliday & Greaves, 2008). Intonation plays a central role in all the languages of the world since it does not only contribute to the flow of discourse, but also to the construction of meaning. So much so that the phonological stratum maps the interpersonal, textual and logical meanings except for the experiential (Halliday & Greaves, 2008). The system of tonality (chunking) and tonicity (placement of accent) maps the textual meanings of language, while the tone system contributes to the mapping of interpersonal and logical meanings. The tone system of English comprises primary and secondary tones which have different functions: primary tones (falling, rising and fall-rising tones) organise the information and contrast different communicative functions, whilst secondary tones, have a role in expressing speaker's attitudes (Halliday & Greaves, 2008;Tench, 1996;Wells, 2006).
Taking this into consideration, it is essential that future teachers of English know, explore and handle the intonation systems of English, so that they can make choices that are meaningful when using the foreign language, and transfer that knowledge to their own teaching practices.

Participants
The data for this research were gathered in two phases (pre -and post-tests), during March and November 2012, with first-year students enrolled at the Teacher Training programme offered by the Language Department (School of Humanities, National University of Rio Cuarto, Cordoba, Argentina). The students (N:50) attending Introduction to Phonetics, the course where the treatment took place, were on average 20 years old and most of them (85%) were attending the course for the first time. Demographic information revealed that 7% of the students had studied English for a number of years and reported having reached an advanced level, 13% of students reported a high intermediate level, 40% reported having an intermediate level after studying for 2-4 years in private language schools and 40% reported a pre-intermediate level after only studying English in high school, where the focus is mainly on the development of reading comprehension skills. The students participating in the study reported here (N:12) were selected on account of the fact that they had attended the instances of pretest and post-test. It is important to mention that since many fail other courses in the first year of the programme, they decide to retake the course where the treatment takes place the following year. This is one of the reasons why many students failed to attend the post-test.

Materials
The data for this study were gathered using pre-and post-tests. The instrument, which was specifically designed and piloted for this research, is divided into three sections: a demographic survey, a perception test and an oral production test. The demographic survey was designed to obtain personal information from each participant, particularly related to age, years of study of English and the level reached. In addition, the survey contains questions that aim to gather information about the participants' perceptions of the aspects of phonetics they find difficult to learn. The perception section includes tasks that aim to assess students' perception of the functions of intonation and their ability to identify tonality, tonicity and tone. The oral production section contains tasks that aim to prompt students to describe a particular text (high-intermediate level) regarding field, tenor and mode to segment it accordingly, and to record themselves reading the text aloud (Di Nardo & Bina, 2015). In the test situation, students are given general information about the research project and they are asked to grant their consent to participate.

Procedures
The treatment was carried out using a set of guides containing awareness-raising activities, practice tasks and reading materials that were specifically compiled for the project (Barbeito et al., 2013). Specific instruction took place during class periods of 2 hours per week, 44 hours altogether during the academic year 2012. Instruction focused on teaching and practicing segmental features such as English vowels and consonants in isolation. In addition, the instruction also focused on supra-segmental features by devoting class time to the reading aloud of different types of genres. During treatment, students were taught to analyse texts from the perspective of the systemic functional theory of language, particularly in relation to field, tenor and mode; and also, to segment texts in relation to tonality and mark them for tonicity and tone before reading aloud.
The same procedures were followed during pre-and post-tests that took place before and after the training. For the recording session, participants were allowed a few minutes to rehearse the text on their own and then read it once to record it. The oral data were recorded using Panasonic RR-US511 digital recorders. The recording of the text was supervised by one of the course instructors.
For the research reported here, the focus will be on students' oral production; however, information gathered from the demographic and perception sections of the tests will also be used to complete the analysis. The oral production sections in the pre-and post-tests were assessed and scored using an ad hoc five-band scale ranging from poor to excellent (Table 1; Cardinali & Di Nardo, 2011). The scale was designed considering the credit requirements established by the National University of Rio Cuarto and takes band 3 (good) as the passing score of the scale. The scale is organised in such a way that students' oral performance is assessed in general terms as well as in relation to specific aspects of the intonation system by means of auditory analysis. Teachers assess how students sound in general terms and then, how they perform taking into account each of the intonation systems at a time.
Students' production of segmental sounds is recorded and graded in a different section of the evaluation sheet.
For this study, two experienced raters listened individually to students' recordings of their reading aloud sections of the pre-and post-tests and rated them separately. Then, the raters met and compared their assessments of students' production and reached agreement on their ratings of tonality, tonicity and tone as well as on the overall assessment of the reading aloud sections. It is important to point out that one of the raters was also the course instructor who carried out the treatment and the other one was an independent rater, specialist in the area of pronunciation and teacher trainer in the Language Department.

Results and discussion
The analysis of students' responses to the questions about register variables of field, tenor and mode in the pre-and post-test shows that 68% of the students provided correct answers in both instances, 16% of the students made mistakes in more than two questions when analysing the register variables during the pre-test and 16% made mistakes in only one question in the analysis during the post-test. It is important to highlight that even though some students made mistakes in the post-test, the number of correct answers was higher. These results seem to suggest that these students manage to identify register variables even though some of them have a limited command of the language upon entrance to the programme; and that a year in the teacher training programme seems to contribute to their understanding of context variables.

Was there improvement in students' overall perfor mance after treatment?
The answer to this question is related to students' overall oral performance which includes an evaluation of the production of segmental features together with tonality, tonicity and tone. The analysis of the data showed that students' overall performance improved in 75% of the cases and remained the same in 25% ( Table 2). The observations made by the raters with regard to students' overall oral performance during the pre-and post-test sessions highlighted the fact that the production of sounds in connected speech affects their performance, in the sense that some of the students who achieved low scores showed difficulty in producing certain sounds of the English inventory, together with the fact that some stretches of connected speech seemed to be particularly challenging. It was observed that many of these students' performances present more than one of the following problems: a) insertion of new words that are not in the script, b) repetition of words, c) insertion of extra sounds and d) omission of words. This is expected since errors in one area may be accompanied by errors in other areas (Pickering, 2000in Pickering, 2009. With regard to the systems of tonality, tonicity and tone, the raters observed that overall performance was mainly affected by mistakes at the level of tonality and tone. At the level of tonality, difficulties were identified when participants could not decide where to chunk certain stretches and at points where they had to stop to guess the pronunciation of an unknown lexical item. At the level of tones, raters observed that the lower scoring participants were those who used a narrow range of tones, mainly primary ones, and failed to produce the secondary tones that might have been expected considering the variables of field, tenor and mode (Halliday & Greaves, 2008;Tench, 1996;Wells, 2006). All these elements led raters to perceive that intelligibility was negatively affected by the tone and key choice (Pickering, 2009). Although some participants showed difficul ties at the level of tonicity, raters considered that these instances did not affect the overall oral performance as much as the other systems described. When considering the scores of the post-test and student's level of English upon entry into the teacher training programme, it is observed that students who have devoted more years to the learning of English; and thus, may be expected to have a good command of the language, tend to obtain scores in the higher bands of the scale; while those who have studied for 4 years or less tend to group in the lower bands of the scale (Table 3). In spite of the fact that the sample in this study is small, we decided to conduct a paired-samples ttest to compare the overall assessment results obtained by the students during the pre-test and the post-test. There was a significant difference in the scores for overall assessment scores in the pre -test (M = 2.08, SD = 1.165) and for overall assessment scores in the post-test (M = 2.83, SD = 1.030) conditions; t(−5.745) = 0.000, (p = 0.05). To begin with, considering the aims of this study, these results are encouraging because they seem to indicate that teaching intonation within the SFL framework does contribute to the development of the oral skills of students, particularly with regard to intonation. However, since these students are exposed to study and produce the target language in other simultaneous courses offered in the teacher training programme, no direct correlation can be assumed.

Which was the aspect that showed most improvement?
The analyses of students' performance in relation to the systems of tonality, tonicity and tone show that students' improvement took place in all the systems, but the system that showed most improvement was the system of tone (42%), while in the systems of tonality and tonicity there was improvement in 33% of the cases. Even though during treatment all the systems received the same focus of instruction and practice throughout the year, it seems that instruction has a positive effect on students' oral performance because after treatment they are more likely to produce a broader range of tones in their reading aloud and this seems to affect the overall assessment of their oral production positively. The results are consistent with the ones reported in Gordon, Darcy and Ewert (2013) who conclude that explicit suprasegmental instruction may lead to improvement as regards comprehensibility.
A comparison of the participants' oral performance with their ability to identify tones (in the perception section of the pre-and post-tests) shows that their improvement in reading aloud as regards the production of tones may be related to the development of the ability to recognise them (Table 4). This may be explained by the fact that during instruction, participants' attention is drawn to the range of tones that are available in English, and that progress in oral production can be related to the awareness raised during the course and the different activities that have been designed for practice. When analysing students' performance during the pre-test in relation to their level of English, it is observed that 83% of the participants get a score of 'poor' and 'fairly good' and that their level of English upon entrance does not make a difference as regards the production of tones. In addition to this, when observing their scores during the post-test, it is observed that only 50% of the participants remain in the bands below 'good'. This may lead us to think that the teaching of the tone system of English may make a difference in students' oral performance even in cases where students may not have a good command over the language. Although the improvement observed cannot be attributed only to the treatment carried out during the year of specific instruction on phonetics and phonology, the results seem to be encouraging. This is all the more promising considering that students in our training college and foreign language users learn the language in classrooms, and in general, do not have the possibility to interact with native speakers or non-native speakers in real communicative events outside the classroom.

Conclusions
The study presented here had the purpose of arguing in favour of the teaching of intonation from the perspective of SFL. To this end, it aimed to answer the following questions: a) Does the teaching of intonation from the perspective of SFL affect the performance of students during reading aloud in an Argentine teacher training university course on phonetics?, b) If so, which system is most affected?
The results seem to indicate that the teaching of intonation within the framework of SFL positively affects the development of students' oral skills, particularly in relation to the systems of tonality, tonicity and tone. Since the effects showed to be more positive in relation to the tone system, it is important to consider making adjustments as regards the instruction of tonality and tonicity systems.
Even though the value of reading aloud as a means for the assessment of intonation may be argued (Ashby & Taniguchi, 2009), the results of this study are considered to be not only positive, but also useful since such a task is of great value for the teaching and training of future non-native teachers of English. There are two important reasons for this: the development of good reading skills is an asset for any student who is at university level; and also, an invaluable tool for future teachers who will have to read not only assignments to their students, but also a variety of genres that may be useful in teaching the foreign language and culture. This is true provided the activity of reading aloud is integrated into a methodology for teaching and assessing that incorporates other, more communicative forms, as well. Considering this, the choice of this specific task may be limiting in the sense that it does not represent what students produce in spontaneous speech or interactions; therefore, further research should be designed towards assessing students' performance in more communicative and interactive activities.
Although as instructors it may be tempting to attribute the improvement to the work done in the course of phonetics and phonology, it is important to consider that the improvement may also be attributed to other factors that are related to the instruction students receive in other courses such as language and grammar, which are part of the programme for first-year students in our teacher training programme.
Much work remains to be done with regard to the role of intonation in communication among nonnative speakers who share the same mother tongue background, which is a characteristic of the context of this study. It is considered that such work will make contributions that may be of value to foreign language teacher trainers in higher education as well as for teachers who offer in-service workshops in different educational contexts.