The Role of a Pronunciation LMOOC in Higher Education Studies

This paper examines the creation and running of the Language MOOC (LMOOC) “The Acquisition of English Pronunciation through Songs and Literary Texts” (1st ed.) and its effects on the proficiency rates of English pronunciation of a group of Spanish speakers registered in the Degree in English Studies offered at the Distance Learning University in Spain (UNED). The LMOOC included a new approach to phonetics teaching and learning based on the stress and rhythmic patterns of literary and music forms. It was offered as voluntary, complementary material to the 640 students registered on the second-year compulsory course on English Pronunciation. The results of a final oral exam showed that the students who took the LMOOC did much better in their oral production than those who only worked with the regular course materials. The LMOOC allowed us to investigate a reversed methodology to phonetics teaching, from rhythmic patterns to sounds, which proved to be highly beneficial to the students of higher education programs with a relevant impact on their pronunciation competences and skills. These findings also suggested that using an implicit methodology to phonetics teaching based on poems and songs is a good complement to explicit learning.


Language Massive Open Online Courses (LMOOCs)
This paper investigates the advantages and disadvantages of using a Massive Open Online Course (MOOC) to supplement learning in higher education courses. In particular, it examines how a Language MOOC (LMOOC) on English pronunciation can help undergraduate students to improve their proficiency rates in their oral production. The LMOOC offers the possibility of 1) testing a new approach to phonetics teaching and 2) studying the implications of combining formal tuition with implicit learning.
MOOCs emerged in 2008 as a new model of online teaching (Downes 2008) coexisting with other forms of distance learning (Baker 2012;Castrillo 2014). MOOCs were conceived as free courses which required no entry qualifications and which were addressed to students of any age, nationality and social background ). This new system of online tuition reached wide audiences and was soon accepted as a valuable source of opportunities at individual, institutional, and social levels (León-Urrutia et al. 2018). Even though one of the initial aims of MOOCs was to open education to a general public, recent research has shown that MOOCs also have a great impact on higher education tuition (Dennis 2012;Yuan & Powell 2013;Iqbal et al. 2014;León-Urrutia et al. 2016, 2018. The wide applicability of MOOCs to both inward (university) and outward (general public) facing students has led researchers such as Czerniewicz et al. (2014) to propose three MOOC categories delineated according to their purposes: 1) formal (credit bearing and tightly curriculated) MOOCs, 2) semi-formal (not part of a degree but may grant credits) and 3) non-formal MOOCs (with no credits and no curriculum alignment).
The MOOC presented in this paper "The Acquisition of English Pronunciation through Songs and Literary Texts" falls into the semi-formal category since it is not exactly integrated into a higher education curriculum but has served as complementary material to the official undergraduate course English Pronunciation as part of the Degree in English Studies held at the Universidad Nacional de Educación a Distancia (UNED) in Spain. The MOOC is classified under the taxonomy of Language Massive Open Online Courses or LMOOCs, a term coined by Bárcena and Martín-Monje (2014) to define those web-based online courses for foreign languages with unrestricted access and potentially unlimited participation.
Up to now, LMOOCs have been used for the teaching of several foreign languages, such as English, French, Italian, Spanish or Chinese, and have covered different skills and language levels (from beginners to advanced) (see Castrillo &Sallam et al. 2020, among others). Despite the great potential of LMOOCs as platforms for foreign language tuition, authors such as Bárcena, Martín-Monje and Read (2015) have also pointed out the difficulties that both language teachers and learners face due to the impersonal nature of MOOCs and their inability to promote social learning. In particular, students cannot communicate with their peers or instructors in the target language. This isolation has a negative impact on the learners' motivation and exacerbates low retention rates and a high level of failure or dropping out (Khalil & Ebner 2014). Despite these drawbacks, studies such as Ventura and Martín-Monje (2016), Whitmer, Schiorring and James (2014) or Vorobyeva (2018) have shown that there are some skills which can be more effectively taught in an LMOOC, such as reading and listening, as opposed to writing and speaking which require more feedback and oral practice.
The LMOOC presented in this paper aims at teaching English pronunciation to Spanish speakers. It focuses on listening and ear-training activities based on the singing and reciting of songs and literary texts. This approach also allows us to examine the effects of implicit learning on the acquisition of English pronunciation by non-native speakers. Whereas explicit learning involves the conscious acquisition of knowledge through formal instruction, implicit learning consists of acquiring knowledge non-intentionally, that is, the learner is not conscious of what he or she is learning (Frensch 1998;Hayes & Broadbent 1988;Reber 1967). Even though for a long time both approaches were presented as categorically different, more recent research has proved that the two processes can co-exist and work simultaneously (Rossetti & Revonsuo 2000;Sun, Slusarz & Terry 2005;and Dörnyei 2009) and that the combination of both methodologies proves to have positive effects on the learning/acquisition process (Latinjak 2014). The LMOOC presented in this paper combines explicit instruction on English phonetics with implicit learning through singing and reciting.
Thus, the goal of this paper is three-fold: 1) to explore the benefits of the LMOOC "The Acquisition of English Pronunciation through Songs and Literary Texts" as a teaching complement to official regulated subjects, 2) to analyze the effects of the LMOOC on the proficiency rates of English pronunciation performed by a group of Spanish speakers registered on the undergraduate course English Pronunciation and 3) to examine the effects of complementing formal tuition in phonetics with an implicit methodology based on songs and literary texts.

Teaching English pronunciation
Teaching English pronunciation to Spanish speakers is not an easy task due to the drastic differences between the two phonetic systems (Estebas-Vilaplana 2009). For example, whereas the vocalic system of standard Spanish consists of five vowels, the English one (as in Southern British English) contains twelve vowels. Furthermore, none of the English vowels coincides in quality with any of the Spanish ones. Similarly, Spanish has 19 consonants as opposed to 24 English consonants. With respect to prosodic features, English and Spanish also differ in many areas, such as in the levels of stress, the rhythmic patterns, the focalization strategies and the intonation patterns. Teaching English pronunciation becomes even harder in a distance learning environment where there is no face-to-face tuition and there is no possibility of giving immediate feedback to the students.
Over the last fifty years, the subject English Pronun ciation or English Phonetics has been part of most degrees in English Philology or English Studies at Spanish universities. The first syllabuses on English pronunciation mainly focused on articulatory phonetics and covered the description of segmental features, that is, how vowels and consonants are produced and combined. More recently, the syllabuses have incorporated suprasegmen tal features (also known as prosodic features), which deal with phenomena that take place at the syllable level such as stress, rhythm, and intonation. According to Jenkins (2000) and Walker (2011), suprasegmental features have proved to be crucial for a good command of English pronunciation. In this paper, a new approach to phonetics teaching is presented, which moves from suprasegmental features to sounds and which is implemented through an LMOOC.

A reversed approach to phonetics teaching and learning
In the year 2018-2019, a team of teachers from the Departamento de Filologías Extranjeras y sus Lingüísti cas at the UNED who are also members of the INME-3L Teaching Innovation Group, decided to create an LMOOC entitled Aprendizaje de la pronunciation inglesa a través de canciones y textos literarios ("The Acquisition of English Pronunciation through Songs and Literary Texts") with two main purposes: 1) to help students of the course English Pronunciation feel more confident about the subject and, in particular, about their production in the oral exam; and 2) to turn English pronunciation into a more enjoyable and friendly subject which could reach a broader audience, as the LMOOC was also addressed to any person interested in the subject. In order to achieve that, a reversed approach to phonetics teaching was adopted. This involved teaching phonetics from prosody to sounds, that is, from the rhythmic pattern of sentences to specific vowel and consonant contrasts. Therefore, the LMOOC follows the opposite approach to most higher education courses on English phonetics, as they usually start with the description of the vocalic and consonantal systems to finally present prosodic features. The implementation of this new methodology was possible due to the flexibility both in terms of content and syllabus design that a MOOC format allows.
This new approach to phonetics teaching was based on the results of several studies on how prosodic features influence the English pronunciation of non-native speakers. The works conducted by Barreiro, Estebas-Vilaplana and Soto (2005), Estebas-Vilaplana, Barreiro and Soto (2006), Arines (2008), and Galguera (2019) showed that Spanish students of English improve their English pronunciation when they recite a poem or sing a song in English, since the rhythmic and the stress patterns of the verse have a positive effect on the pronunciation of the overall text.
The main reason why Spanish students improve their English pronunciation when they sing or recite is closely related to the differences in the rhythmic patterns of the two languages. While Spanish is a syllabletimed lan guage, that is, is a language in which syllables (no matter whether they are stressed or unstressed) tend to be produced at equal time intervals, English is a stresstimed language, which involves a similar amount of time from one stressed syllable to the next (see Cruttenden 2008; and Roach 2009, among others). Thus, the higher the number of unstressed syllables between stresses in English, the quicker they will have to be uttered so as to maintain a similar amount of time between stressed syllables. This has an effect on the production of vowels. In order to be able to compress the duration of syllables between stresses and keep a stress-timed rhythmic pattern in English, most unstressed syllables tend to be produced with a weak vowel (see Estebas-Vilaplana 2009. Therefore, the pronunciation of vocalic sounds in English highly depends on the stress condition of the syllable. Rarely is this idea presented in a class of English as a second language. Thus, the English pronunciation of most Spanish speakers is based on the syllable-timed rhythm typical of Spanish. The immediate consequence of this practice is that Spanish speakers produce all English syllables with an equal length, with a full vowel quality and without any vowel weakening, as in Spanish. Based on this evidence, the LMOOC included songs and literary texts (both poems and prose) as the starting point to teach English pronunciation. Songs and literary texts, especially those written in verse, create certain restrictions on the production of sentences which have to accommodate to the rhythmic structure. Thus, the LMOOC was designed with the idea that the rhythmic patterns of the songs and poems would help students to focus on the production of the vowels in the stressed syllables and subsequently weaken the production of the vowels in the unstressed syllables. Once the students had learned this process, they could apply it to prose and finally to spontaneous speech.

Description and analysis of the LMOOC
The LMOOC "The Acquisition of English Pronunciation through Songs and Literary Texts" ran for four weeks, in two editions: from March 11 to April 7, 2019, and from March 9 to April 6, 2020. The course was divided into four modules with the following content: Module 1 (English Rhythm); Module 2 (English Vowels); Module 3 (English Consonants), and Module 4 (Conclusions). Modules 1 to 3 were further subdivided into five sections, each of them corresponding to short portions of content presented in video-recorded lectures. Each video recording lasted an average of 15 minutes. Overall, the course consisted of 16 video-recorded lectures. The progress of the participants was assessed by a battery of 20 multiple-choice perception exercises with immediate feedback included at the end of each module. Besides, there were discussion forums where participants could interact and pose questions and comments to their peers and also to the teaching team. The course also included an initial and a final questionnaire to gather information about the participants and find out whether the course content had met their expectations.
1,709 students were registered for the first edition and 4,160 for the second edition with the median learner age being 44 and 40 respectively. In both editions more than half of the participants were aged over 40 years (see Figure 1).
As far as the level of education of the participants, measured in terms of completed studies, more than half held a university degree and almost a quarter had an MA or a PhD (see Figure 2).
With respect to gender, 70.2% and 68.3% of the participants were female and 28.5% and 30.4% were male in the two editions. Finally, according to geographic data, people from 14 countries registered for the course. Most of the MOOC participants were from Spain (86%/79% in the two editions) but there were also a few learners from Latin American countries (Colombia, Costa Rica, Mexico, Bolivia, Ecuador, Guatemala, and Venezuela) and  European countries, such as Bulgaria, Germany, UK, Italy, Luxemburg, and Portugal.
The LMOOC received very good reviews from the participants. According to the students who answered the final questionnaire (469/557), 92% and 97% indicated that the course had accomplished their expectations in terms of English pronunciation learning (see Figure 3).
Furthermore, 97.2% and 98.3 of the participants who answered the questionnaire maintained that the LMOOC had helped them to improve their English pronunciation (see Figure 4).
Finally, as presented in Figure 5, the participants believed that the most useful topic presented in the course had been the module on rhythm (26,2% and 32%) while 64%/62% of the participants in both editions answered that all topics were useful.
These results revealed that the new approach to English phonetics teaching had a positive effect on the students of both editions of the course who found it useful to improve their pronunciation skills. A crucial observation is that most of the participants found the module of rhythm the most useful. This is closely related to the fact that suprasegmental features are hardly ever covered in an English class and hence the students appreciated tackling English pronunciation at the sentence level with the implicit help of music and literary forms.

Experimental Design
Although the LMOOC was offered to a general audience, this section analyses the effects of the LMOOC on the pronunciation skills of those students registered on the undergraduate course English Pronunciation as part of the Degree in English Studies at the UNED. English Pronuncia tion is a second-year course with 150 hours of online tuition which lasts four months. Its main aim is to provide guidelines, tools and methodological resources for the autonomous learning of English pronunciation. Students have access to a virtual learning platform which contains all the course materials as well as forums for peer discussion. The syllabus covers the typical content of an  English pronunciation subject offered in higher education courses. The topics are presented in the canonical order, that is, the course starts with vowels, consonants and combinations of sounds, and it finishes with prosodic features, namely, stress, rhythm, focus and intonation. At the end of the year, the pronunciation competences of the students are measured by means of two tests: 1) a written exam which covers some theoretical issues and a phonetic transcription, and 2) an oral exam which assesses a variety of pronunciation phenomena, including both segmental and prosodic features, as well as the reading of a paragraph in phonetic transcription. Since the first time this course was offered in 2009, one of the main concerns of the students has been how to approach the oral part and succeed in the oral exam. Despite the great number of supportive materials provided in the course to do the oral test, such as video-recorded lectures, a mock exam and well-defined assessment criteria, including evaluation rubrics, the vast majority of students feel worried about this test and how to succeed in it. The incorporation of the LMOOC as a voluntary, extracurricular activity to the course aimed at giving students more confidence in their oral exam.
Overall, 142 students (out of 640) who registered in the English Pronunciation course in the academic year 2018-2019 completed the first edition of the LMOOC. The effects of the LMOOC on the pronunciation skills of these students were assessed by means of the final oral exam. It was expected that relevant differences would be found in the pronunciation of the students who took the LMOOC as opposed to those who did not. The two working hypotheses were as follows: 1. The students who complete the LMOOC will show better proficiency rates in English pronunciation in the oral test than those who do not complete it.  Estebas-Vilaplana and Solans: The Role of a Pronunciation LMOOC in Higher Education Studies Art. 21, page 6 of 10 2. The students who complete the LMOOC will do particularly better in the production of a long paragraph (in phonetic transcription) since it focusses on the pronunciation of sentences rather than specific sounds.

Informants
The students who took part in this study were those who sat for the oral exam of the subject English Pronunciation in the academic year 2018-2019. Table 1 presents the number of students registered on the course as well as those who did the oral test, with or without having completed the LMOOC. As exhibited in Table 1, 334 students out of 640 did the oral test. The number of students who completed the LMOOC and sat for the oral exam was 142 out 334. Finally, 192 students did the oral test but not the LMOOC. Thus, group A (students with LMOOC) is made up of 142 informants and group B (students without LMOOC) includes 192 informants.
Both groups of students followed the same tuition guidelines and worked with the same types of teaching tools, namely, video-lectures, a textbook and a software package with ear-training materials and exercises. The type of tuition offered in the course was mainly explicit, with formal instruction on each topic. Additionally, the students who voluntarily joined the LMOOC (Group A) could also benefit from an implicit learning methodology since they could approach English pronunciation through songs and literary texts, with a less formal attention to phonetic issues.

Data gathering
The data to assess whether the students who had completed the LMOOC showed better proficiency rates in English pronunciation, was gathered as part of the oral exam of the subject English Pronunciation in the academic year 2018-2019. The oral test is recorded by means of the eoral application, which is a software designed at the UNED for doing oral exams on-line Alba Juez et al. 2006). The oral test is divided into three parts: 1) the production of English vowels, consonants and consonant clusters, 2) the production of English stress, rhythm, focus and intonation patterns, and 3) the reading of a phonetic transcription which assesses the pronunciation competences at the discourse level and covers fluency, intelligibility, weak forms and connected speech processes.

Results
The results of the oral test showed that those students who had completed the LMOOC obtained better proficiency rates than those who had not done it. Figure 6 shows the percentage of students who passed the oral test with and without completing the LMOOC. As exhibited in Figure 6, 100% of the students who completed the LMOOC succeeded in the oral exam. The percentage of students who passed the oral exam without LMOOC is 67%. A t-test comparing the results of the two groups highlighted significant differences at p < 0.05. These results confirm that the LMOOC has had a positive effect on the success of the oral test.
These findings are corroborated by the final marks obtained by the students in the oral exam. Figure 7 shows the percentage of students according to their final marks (i.e. between 10 and 9, 8 and 7, 6 and 5 or less than 5, that is, a fail). As observed in Figure 7, the students who took the MOOC had better grades than the students who did not complete it. 34% of the students with LMOOC obtained a mark that ranged between 10-9, 45% between 8-7, and   Table 1: Number of students who completed the LMOOC and did the oral exam.

Number of students in the course English Pronunciation in 2018-2019
Overall number of students in the course English Pronunciation 640 Number of students who did the oral test 334 Number of students who completed the LMOOC and did the oral test 142 (Group A) Number of students who did not complete the LMOOC but did the oral test 192 (Group B) 21% between 6-5. None of the students who completed the LMOOC had a fail in the oral test. The students who did not complete the LMOOC had worse results, with 14% of the students with 10-9 final marks, 25% with 8-7, 28% with 6-5 and 33% with a grade lower than 5. The final mark mean value for those students who completed the LMOOC is 8.1 (out of 10), as opposed to 5.7 (out of 10) for those students who did not take the LMOOC. The results of a t-test showed that the differences in the oral exam grades for the two groups are significant at p <= 0.05. Thus, the results relating both to the percentage of students who passed the oral exam and the percentage of their final marks show that the students who took the LMOOC did much better in the oral exam than those who did not take it. This corroborates our initial hypothesis that the LMOOC has a positive effect on the proficiency rates of students of English pronunciation. This further indicates that the incorporation of a MOOC as extracurricular input to undergraduate courses can be highly beneficial for the students.
Finally, Figure 8 shows the percentage of correct responses in the three tasks included in the oral exam for the students with LMOOC and those without LMOOC. The first task, relating to the production of consonants and vowels, does not show significant differences (p > 0.05) in the production of the two groups (89% of correct responses for the students with LMOOC and 85% for the students without LMOOC). This is a rather expected result since the practice of vowel and consonant contrasts is recurrent in the course materials. With respect to the production of prosodic features, the results show a higher number of correct responses for the students who completed the LMOOC (76%) as opposed to those who did not do it (61% of correct responses). The results of a t-test show that the differences between the two groups are significant (p < 0.05). This suggests that students could better grasp the rhythmic and  accentual patterns of English by paying attention to texts with a well-defined rhythmic structure, such as music and literary forms. This also indicates that combining explicit and implicit learning can be beneficial for the acquisition of English pronunciation.
The final task of the oral exam consisted of reading a passage in phonetic transcription. This is the task that tends to be more difficult for the students because it covers the pronunciation of sentences rather than concrete sounds. As presented in Figure 8, the LMOOC had a positive effect on the production of passages in phonetic transcription, since the students who had completed the LMOOC exhibited 24% better results than those who did not do it (54% vs. 78%). The results of a t-test showed significant differences between the two groups (p < 0.05). These results confirm the second hypothesis that the LMOOC had a greater impact on the expected pronunciation of sentences and paragraphs than on the actual production of single sounds. Thus, approaching English pronunciation from the rhythmic pattern of sentences has proved to be a suitable method for helping students tackle the production of longer texts.

Discussion
This study has shown how an LMOOC on English pronunciation and literature has helped the undergraduate students of the course English Pronunciation improve their oral production skills. This research has theoretical implications on 1) the implementation of a reversed methodology to phonetics teaching, 2) the usage of MOOCs to support registered undergraduate students as opposed to the general public, and 3) the combination of implicit and explicit approaches to language learning.
One of the main advantages of the LMOOC was that it offered the opportunity to test a new method of English pronunciation teaching, from rhythm to sounds. This reversed approach to phonetics teaching, firstly proposed by Barreiro, Estebas-Vilaplana and Soto (2005) and Estebas-Vilaplana, Barreiro and Soto (2006), explains the phonetic system of English and, more specifically, the production of vowels, in relation to the stressed-timed rhythmic pattern of sentences. This new approach could not be easily included in the syllabus of a regular higher education course since the contents are well-established by the degree plan and modules. However, the LMOOC format allowed us to develop a different syllabus organization which involved phonetics teaching from prosody to sounds and which proved to be highly beneficial for Spanish students of English. It would be very interesting to further explore the integration of this approach to the actual course curriculum in order to examine the real impact of a reversed methodology to English phonetics teaching. This new approach can also be applied to other languages, especially those with a syllable-timed rhythmic structure.
The results of this study also support the claims of Dennis (2012), Yuan and Powell (2013), Iqbal et al. (2014), Czerniewicz et al. (2014), Schuwer et al. (2015) and León-Urrutia et al. (2018), among others, that MOOCs are not only favourable to the general public but also to students of higher education courses. This study has shown that the students of the undergraduate course English Pronunciation who successfully completed the LMOOC showed better results in their oral production than those who did not take the LMOOC and only worked with the regular course materials. However, this research has some limitations. First of all, there was no control on the motivation and dedication of the students who took the LMOOC as opposed to those who did not join it. Thus, it is possible that the LMOOC students were more devoted than the others and that their good results in the oral test should be associated not only to the effects of the LMOOC but also to their levels of motivation. Second, no baseline tests were provided before the LMOOC, and hence the good results of the students could also be related to their proficiency levels at the beginning of the course. In future editions of the LMOOC, initial tests to determine the proficiency levels of the students and their motivation will be included.
The results of the LMOOC also had some theoretical implications on the dichotomy between implicit and explicit learning. As suggested in many works such as Rossetti and Revonsuo (2000), Sun, Slusarz and Terry (2005), or Latinjak (2014), among others, the combination of the two approaches is highly beneficial for students of foreign languages, and in particular of pronunciation (see Rubio 2014). In the regular university courses, English pronunciation is mainly taught following a formal, explicit approach. The LMOOC has allowed us to incorporate songs and literary texts so that students could grasp pronunciation in a more informal way. The impact of this new approach on phonetics teaching needs to be explored among students of other degrees which do not belong to language studies, in order to see if the new methodology can also be beneficial to students with little background on linguistic issues.

Conclusions
This paper has examined how an LMOOC can be used as a supplement to traditional higher education courses. In particular, it has analysed the effects of the first edition of the LMOOC "The Acquisition of English Pronunciation through Songs and Literary Texts" on the production skills of the students registered on the undergraduate course English Pronunciation from the Degree in English Stud ies (UNED). The LMOOC proposed a new methodology for English phonetics teaching which adopted a reversed approach, from suprasegmental features (rhythm and stress) to sounds (vowels and consonants), and an implicit tuition based on poems and songs. This new approach to phonetics teaching and learning has been highly accepted by the LMOOC participants and proved to enhance their English pronunciation skills. The results of this study confirm that MOOCs hold huge potential as complements to official regulated courses in higher education.

Funding Information
This research is currently being undertaken with the support of Universidad Nacional de Educación a Distancia (UNED) in the context of the Teaching Innovation Group