Conceptualising language MOOC diversity: The creation of a defined taxonomy through the analysis of course indicators

Abstract

Language MOOC research has experienced a notable evolution from practice to conceptuality since its emergence as a subdiscipline of computer-assisted language learning. The versatility of the MOOC format for language learning has led to experimental designs that combine linguistic acquisition with other educational activities. This has been considered to be conducive to new ways of understanding how language learning occurs in LMOOCs, although there is no solid classification of LMOOCs subtypes to date based on course design. This study aimed to contribute to the conceptualisation of the field by creating a taxonomy for existing LMOOCs. Grounded theory strategies were adopted, so evidence was systematically collected to develop conceptual categories based on a thorough analysis process of the syllabus and short description of 432 courses. As a result, six LMOOC modalities emerged from the analysis: general language learning LMOOCs, LMOOCs for academic purposes, LMOOCs for professional purposes, LMOOCs focused on a specific language skill development, cultural-oriented LMOOCs, and meta-language learning LMOOCs. This study means a significant contribution to the LMOOC research field inasmuch as it is one of the first empirical-based attempts to broaden the definition of LMOOC.

1. Introduction

One of today’s widespread online modalities for foreign language learning are language MOOCs (LMOOCs), which are “dedicated Web-based online courses for second languages with unrestricted access and potentially unlimited participation” (Bárcena & Martín-Monje, 2014: 1). LMOOCs are considered nowadays an innovative educational format based on the open and social side of language learning in online environments (Martín-Monje, 2023).

Early research has already suggested that LMOOCs need to foster real language communication and practice, provide interactive and motivational educational materials, and explore the cultural nuance of the target language as a complement to it (Perifanou & Economides, 2014). Those aspects are built up on components of language learning such as passive assimilation of rules and vocabulary, but more prominently on acquisition of skills and their application in practice (Bárcena & Martín-Monje, 2014).

These needs led to the first review of LMOOCs (Perifanou & Economides, 2014) to comprehend the levels of interactivity and, consequently, of effectiveness for language learning by looking at the content, pedagogy, assessment, community, technical infrastructure, and financial issues. The results showed that LMOOCs of that time followed a cognitive behavioural approach. The authors’ classification allowed the identification of language-oriented MOOCs for the first time when LMOOCs were only an emerging tendency.

This discipline has evolved since then. The number of LMOOC users has been progressively increasing, with a rapid acceleration during the COVID-19 outbreak (Martín-Monje & Borthwick, 2021), and research has reached a mature stage where conceptual contributions are more commonly reflected in the literature (Díez-Arcón & Martín-Monje, 2023). Designs are also changing in attempts to find suitable, effective, and successful deliverable modes by combining linguistic knowledge acquisition with other educational activities.

Such experimentation has resulted in LMOOCs that also work on the development of metacognitive strategies (McLoughlin & Magnoni, 2017) and of self-determined learning characteristics (Agonács, Matos, Bartalesi-Graf & O’Steen, 2020), among others. These design proposals, which have been scarcely explored to date, open new ways of understanding and describing LMOOCs that may be conducive to broadening its current conceptualisation by the elaboration of potential classifications for LMOOC typologies.

The widely accepted definition for LMOOCs (Bárcena & Martín-Monje, 2014) was coined with their emergence as courses with distinctive characteristics in relation to other disciplines. This definition, however, should be further developed, and it needs robust scientific contributions that would make the creation of a common conceptual framework possible.

The main objective of this study was to contribute to the existing conceptualisation of the typologies in LMOOCs by complementing their definition. With this aim, a dedicated taxonomy was developed looking at the diversity of LMOOCs. It was based on the systematic analysis of the content (Krippendorff, 2018; Porta & Silva, 2003) of specific course information selected as the main core of meaning subject of study (i.e. brief description and syllabus) (Liyanagunawardena, Lundqvist, Mitchell, Warburton & Williams, 2019). This approach can be considered the first comprehensive review from the perspective of course design.

To this end, empirical data were obtained from Class Central database and were inductively analysed by adopting the constructivist grounded theory (GT) (Charmaz, 2008) approach, which often produces grounded descriptions instead of abstract concepts. The analysis of course indicators aimed to develop and describe new or refined concepts out of the data analysed. The analysis first needed to determine which courses could be regarded as LMOOCs according to the Common European Framework of Reference for Languages (CEFR) description of competences. Later, Charmaz’s constructivist GT procedures were applied, which contain generic GT techniques from other GT schools of thought (Kenny & Fourie, 2015) to analyse and summarise the characteristics of the previously selected LMOOCs. Finally, a robust conceptual taxonomy based on the identified common patterns in the design of LMOOCs was generated.

2. Theoretical framework

The first part of this section explores theoretical references for the modalities employed in foreign language learning and teaching. Next, what learning a foreign language entails and how it is currently understood is explained, with special attention given to the cultural element in language learning and its consistency with the postulates of the communicative approach as considered in the CEFR. Language learning in MOOCs is also analysed to understand what components have been considered to fit into this modality, and, finally, previous LMOOCs categorisations are evidenced to explain how these proposals support the conceptual construction in the discipline.

2.1 Language learning distinctions

2.1.1 General language learning

General language learning can be associated with the CEFR as it encompasses the set of activities and strategies to be developed by language learners. It is not geared towards specific skills, topics, or target groups, but is more general, both in terms of skills development and the topics to be covered. That is the reason why regular language didactic programmes can usually be framed within this distinction (Grapin, 2017). Since the CEFR assumes the communicative approach (Council of Europe, 2020), it is understood that the contents embrace different features of the linguistic, sociolinguistic, and pragmatic competences, leading to an effective communication and stimulating language use on a general level.

2.1.2 Language for specific purposes

Another widespread modality of foreign language learning is language for specific purposes (LSP). It arises from a practical need to assess individual skills in performing specific tasks in academic and professional settings. Traditionally, two streams have been identified within LSP – English for Academic Purposes and English for Occupational Purposes – due to the hegemony of the English language in technological, scientific, and commercial development (Lesiak-Bielawska, 2015). According to Dudley-Evans and St. John (1998), the academic variant relates to the needs detected in the study of a given area, while the occupational variant deals with on-the-job training needs.

LSP has absolute and variable characteristics (Johns & Dudley-Evans, 1991; Strevens, 1977). Absolute characteristics represent (1) the need to meet learners’ needs; (2) the provision of content related to specific disciplines, activities, and occupations; and (3) the adaptation of language at all linguistic levels to the discipline to which it is related. Variable characteristics, instead, include the development of specific language skills.

2.1.3 English as a medium for instruction, content language integrated learning, and content-based instruction

These distinctions use the foreign language as a language of instruction. Foreign language is primarily English because of its status as lingua franca, although it is not limited to it. There are different approaches: English as a medium for instruction (EMI), content and language integrated learning (CLIL) and content-based instruction (CBI), which differ “… on the relative positions of language and content in learning objectives and assessments” (Brown & Bradford, 2017: 328), which permits observation where language learning is a priority.

EMI does not have as its main objective to improve students’ language skills; rather, the focus is on subject-content mastery (Brown & Bradford, 2017). In contrast, CLIL involves a joint learning practice of the subject matter and the foreign language (Smit & Dafouz, 2012), so it is understood to have a dual focus where equal attention is paid to both topic and language.

The definition of CBI is less clear and has opposite views. According to Brown and Bradford (2017), some authors indicate that CBI principally entails the development of academic language skills where the content is merely a vehicle to support students to master the language. Another stream claims instead that the learning of academic subject matter and second language skills are equally considered, overlapping with the distinctive features of CLIL.

2.2 The CEFR

The CEFR (Council of Europe, 2001) is an indispensable frame of reference in all aspects of foreign language education (Little & Figueras, 2022) and it has an influence on all other rigorous initiatives to build common language reference standards. This is the case in Asia, where it is widely used (Higuchi, 2012; Read, 2019), and in the USA, where the CEFR has served as a basis for unifying assessment systems within the Teaching of Foreign Languages Framework of the American Council (American Council on the Teaching of Foreign Languages, 2016).

Thanks to the CEFR, there is a common understanding of how a foreign language should be taught and learnt nowadays. This framework points out that language users are social agents and the language used is an instrument for communication rather than an object to be studied (Council of Europe, 2020).

From this perspective, the individual needs to develop a range of competences to achieve a successful communication in line with the well-accepted socio-constructivist and sociocultural theories for language learning (Ellis & Larsen-Freeman, 2006; Savignon, 2018). The CEFR is also employed to identify a certain competence level (Krajcso, 2016) and to assess the quality of language production using descriptors “… to help to align the curriculum, teaching and assessment” (Council of Europe, 2020: 42)

Communicative language competences, according to the CEFR, are

• linguistic, which refers to the language in use and resources, and knowledge of the language as a system (linguistic range, vocabulary, grammar, phonology, orthography, etc.)

• sociolinguistic, which refers to the knowledge and skills required to deal with the social dimension of the language (politeness conventions, humour, the use of idioms, etc.)

• pragmatic, which refers to the actual use of the language focusing on the discourse, structures, and functions (flexibility, turn-taking, coherence and cohesion, fluency, etc.)

• plurilingual and pluricultural, which refers to the use of all the linguistic resources to see similarities, regularities and differences between languages and cultures (switch from one language or dialect to another, express oneself in one language and understand a person speaking another, etc.).

2.3 The importance of culture in the communicative approach for language learning

There is an inherent relationship between culture and language (Byram & Wagner, 2018; Choudhury, 2014; Marhamah, Daud & Samad, 2017), and this connection plays a relevant role in the learning and teaching of a foreign language at various levels (McKay, 2004). It influences the linguistic and pedagogical aspects of the selection of the content or the methodology itself, which makes it difficult to adapt programmes aimed at the acquisition of linguistic competences to the treatment of cultural issues (Halbach, 2002; Stern, Allen & Harley, 1992).

There are different ways to conceptualise culture when applied to language teaching (Byram & Morgan, 1994) depending on the source of their origins: humanistic (Culture, “C”) or anthropological (culture, “c”). The former indicates a lofty approach referring to the concept of culture historically associated with art, music, literature, politics, etc., “as a way of ensuring the continuity of a national community by giving it meaning and value” (Kramsch, 2013: 65). The latter introduces elements of behaviour, thought, way of life, and attitudes and assumptions.

According to Thanasoulas (2001), the cultural element can be better integrated in the language curricula when methods focusing on communicative acts are used, compared to other approaches that do not emphasise such scenarios. Based on this paradigm, the most relevant elements to be integrated in the cultural-linguistic syllabus are (1) factual cultural information about society, geography, and history, among others; (2) cultural behaviour or way of life expressed in language formulas applied to conversation and kinesthetics; and (3) cultural achievements, mainly associated with artistic and literary milestones (Stern et al., 1992).

2.4 Language learning and MOOCs

The MOOC format has been widely used for general language learning and it has also been proposed for LSP since it can fulfil students’ needs of language skills acquisition when studying abroad and be a complement or an extension to university subjects aiming to develop professional language learning (Castrillo & Martín-Monje, 2018; de Waard & Demeulenaere, 2017; Troncarelli & Villarini, 2017; Zhang, 2017).

Looking at the cultural element, computer-assisted language learning has considered it a relevant aspect since its very beginning, although LMOOCs still lack its inclusion in resources and tasks (Wang-Szilas & Bellasen, 2017). The latter authors propose a model to overcome this deficiency by introducing, at the end of each topic, one or two cultural elements in line with the content of the lesson itself. Other authors have combined linguistic and cultural knowledge exploring mythology, contextual knowledge in the linguistic syllabus (coding, register, intonation, etc.), local education systems, or intercultural differences (Mac Lochlainn, Nic Giolla Mhichíl, Beirne & Brown, 2020; McLoughlin & Magnoni, 2017; Rovira-Collado, López, Baile-López & Rivas, 2019).

The LMOOC format and its diversity in terms of participants has the potential to facilitate the understanding of culture, values, and beliefs of one’s own and target language and the training to be able to share and construct knowledge with people from different backgrounds from both majority and minority languages (Bax, 2018; Escolano-López, Leal-Rivas & Rovira-Collado, 2018; Godwin-Jones, 2015; Gollin-Kies, Hall & Moore, 2016; Mac Lochlainn et al., 2020; O’Dowd, 2013; Šmilauer, Pognan & Vigent, 2018).

There is also considerable evidence of experiences with the foreign language as a medium of instruction in MOOCs. Most of the studies showcase teachers’ continuing professional development courses (Cerveró-Carrascosa, 2022; King, Luan & Lopes, 2018; Rutkauskiene, Volodzkaite, Hansen, Murray & Kubiliunas, 2020; Zubkov, 2019), so the focus is on pedagogical issues rather than linguistic ones. However, some authors understand that MOOCs for foreign language teachers do improve foreign language skills by the learning of the profession’s own meta-language (Ardavani, 2020; Mangenot & Phoungsub, 2018; Orsini-Jones, Zou, Hu & Wei, 2017; Viswanathan, 2012).

More modestly, MOOCs using the foreign language for instruction are also found (Read, Sedano & Barcena, 2018; Tanaka-Ellis & Sekiguchi, 2019). It is the elements (content vs. language) to be assessed that determine their possible eligibility as LMOOCs. Experimentation in the massive format has made it possible to combine metacognitive strategies with linguistic acquisition (see “MOVE-ME” ¹ European project). These strategies were provided for good language performance and guidance during the course (McLoughlin & Magnoni, 2017).

The authors argue that although they may not be considered foreign language courses, they should be classified as such, as they help to improve transversal, lexical, syntactic, and paralinguistic skills. This course exemplifies the application of metacognitive strategies for linguistic competences acquisition. As in the case of language teaching courses, further research based on sound classification criteria needs to be carried out to confirm their status as LMOOCs.

2.5 (L)MOOC taxonomies: Contributing to the conceptualisation of the field

There are only two LMOOC classifications to date (Martín-Monje, 2023; Perifanou & Economides, 2014). The earliest one was focused exclusively on the identification of language learning courses based on the interactivity feature because of the needs of an incipient field. Instead, Martín-Monje (2023) approached this topic by analysing the types of LMOOCs dedicated to English learning (ESL/EFL).

Perifanou and Economides (2014) intended to comprehend the levels of interactivity in LMOOCs to determine effectiveness in language learning. They looked at the content, pedagogy, assessment, community, technical infrastructure, and financial issues to conclude that the LMOOCs reviewed did not offer interactive settings and still followed traditional ways of learning languages following cognitive behavioural pedagogical models. The need to promote interactivity brought the approach closer to socio-constructivist principles that have finally been established as learning standards in the field (Díez-Arcón & Martín-Monje, 2023; Sallam, Martín-Monje & Li, 2022).

Martín-Monje (2023) proposed a categorisation of LMOOCs focused on English learning by examining the content of the courses. LMOOCs were classified as follows: (1) general MOOCs, (2) LMOOCs focusing on certain skills, (3) LMOOCs on English for specific purposes, (4) tandem MOOCs combining L1 and L2 (first language and second language), (5) LMOOCs on cultural aspects in English-speaking countries, (6) LMOOCs to prepare for standardised tests; (7) LMOOCs for social inclusion, (8) MOOCs on EMI and CLIL, and (9) MOOCs on language learning methodology.

This categorisation can be considered a first approach to the objectives of the present research, since it considers typologies of LMOOCs, although from a practical perspective. The approach intended in this study differs, therefore, from the mentioned LMOOC categorisations by directly targeting the conceptualisation of the typologies of language learning in MOOCs. Conceptualisation is intended to explain the identification of patterns in an area, which entails understanding the approaches to language learning in this format and contributing to complementing the defining characteristics of LMOOCs.

3. Methodology

This study was carried out between September and December (2022). The conceptualisation of LMOOC typologies in this study needed to be solidly grounded to propose a reference framework for future research. The GT methodology adopted understands conceptualisation as its core category and explains complex realities by establishing their most important characteristics through constant comparison of sampled data. GT is an inductive methodology that provides systematic guidelines for gathering, synthesising, analysing, and conceptualising qualitative data (Charmaz, 2012). This way, research generates an explanation of the facts through the development of concepts leading to refined or new theories.

This study adopted the three major strategies of the GT method, namely coding, memo-making, and theoretical sampling (Charmaz, 2012). These aspects converge in all three factions or historical perspectives for GT (classic, Straussian, and constructivist), and, according to Kenny and Fourie (2015), they are based on the application of the constant comparative technique where “1) Codes are compared with codes, 2) Codes are compared with emerging categories, … 3) Categories are compared with one another …; and 4) The emerging theory is compared with the literature” (p. 1271). Nevertheless, the strategies applied in this study followed the constructivist approach developed by Charmaz (2008), which allows for more flexibility in the interpretation of the coding phases and uses literature at every stage of the procedure.

Each of the strategies were aligned with the main purposes of this study, which first needed to identify the courses that could be classified as LMOOCs to consequently analyse and synthesise the qualitative data for conceptual construction (see Figure 1).

Figure 1. Five-step methodological process applying grounded theory.

3.1 Sample selection

The selection of MOOCs was made through Class Central’s database as a reference platform, which aggregates MOOCs from many of the most relevant providers, such as Coursera, EdX, Udacity and FutureLearn, and other smaller platforms and independent MOOC providers. Data were retrieved from Class Central in September 2022. A first screening was carried out by the selection of the so-called “parent categories” (primary categories defined by Class Central), consistent with their association to language learning (Humanities, and Education & Teaching). Subsequently, the “child categories” (subcategories defined by Class Central) were selected through the review of the titles of the courses included in each subcategory (n = 135). Initially, 12 child categories were selected (see Table 1), with a total of 1,992 courses potentially related to language learning (see Supplementary Materials 1).

Table 1. Data screening in Class Central

3.2 Data screening

Analysis of content technique was used to select the units of analysis or course indicators that constitute the cores of meaning to be studied (Porta & Silva, 2003) prior to the application of the GT strategies. The syllabus and the brief description of each course were the corpus used to identify the units of analysis chosen as indicators, such as words or phrases with relevant information for the purposes of the study. These elements were related to the subject of the course itself, as identified by Liyanagunawardena et al.’s (2019) MOOC taxonomy.

The review of the selection of course indicators first aimed to look at which of the courses could be classified as LMOOCs by (a) identifying if any of the communicative competences defined in the CEFR were mentioned; and (b) if they complied with the basic features of the MOOC format, such as unlimited participation, availability of open educational resources and being free of charge for participants (Castrillo, Martín-Monje & Vázquez-Cano, 2018). These selection criteria resulted in a final sample of 432 LMOOCs, which were analysed by applying GT strategies, as detailed as follows.

It must be noted that the authors manually accessed all the selected courses through the links included in the original dataset provided by Class Central to constantly compare and analyse the syllabus and course description of the selected LMOOCs. The authors created a new dataset using Microsoft Forms (Supplementary Materials 2) and manually entered metadata from the courses. The questionnaire contained relevant information for the online localisation of the courses such as the ID of every LMOOC as classified on Class Central, title, university, parent and child categories, and language in which the course was delivered. Also, it progressively included information out of the different phases described below, such as the communicative competence(s) worked (to determine whether they could be regarded as LMOOCs or not), the labelling indicators (coding stage), and the tentative categorisation out of the description of the codes (memo-writing).

3.3 Coding and memo-writing

The coding phase, an initial strategy envisaged in the GT, overlapped with the identification of LMOOCs, which allowed for a first review of course indicators, as referred to previously. The coding process entailed the identification of recurrent elements of the corpus that were abstracted and labelled (language skill(s) worked, target audience, learning objectives, reference to language levels, content-wise information, etc.) to classify the specific information in each LMOOC. The comparison of different indicators enabled connections to be progressively made between codes based on the common and distinctive patterns encountered.

This second coding stage (focused coding) permitted an outline of tentative categories based on the codes’ recurrence and significance. This second codification set the following temporary theoretical categories: general language learning (270 courses = 62.5%), cultural (26 courses = 6%), language for specific purposes (124 courses = 18.7%), language meta-learning (10 courses = 2.3%), and content-based instruction (2 courses = 0.5%).

This initial categorisation was accompanied by the application of memo-writing. This strategy consisted of describing the codes created, analysing their properties, and specifying the conditions in which they occurred. Categories, therefore, were complemented by specific information of interest found by the reviewers (e.g. “Language skills acquired by cultural issues (syllabus)”, “Just writing-focused practice”, “Tips, tricks, how to improve?”, “Discover British culture while improving the English language”, “Basic level French/routinary activities explored”, “Study of the most common mistakes of German writers”, etc.). This stage, along with the former analyses, enabled the first possible definitions for the codes to be shaped and the data to be better interpreted (category creation) by demonstrating the existing relationships among the distinctive categories of LMOOCs according to their synthesised common features.

3.4 Theoretical sampling

Theoretical sampling entails developing and/or defining emerging theoretical categories based on the core theoretical categories created in the previous stages. It needs more in-depth checking of the properties, boundaries, causes, and consequences of these theoretical categories. It is assumed that these actions allow for building precision, density, and complexity into the emerging theoretical statements by inductive and deductive reasoning (Charmaz, 2001).

To this end, the definitions outlined in previous phases were reviewed and complemented based on the determination of the relevance and adequacy of the theorised concepts composing the sample. Thus, robust and justified categories were established and allowed the presentation of the taxonomies proposed to define different approaches to language learning in MOOCs based on course design.

This final output was compared with existing literature, and it was ascertained that the emerging concepts from this research partially matched in nature with some of the theoretical approaches to language learning retrieved from the literature, although new concepts were also developed and existing modalities for language learning were refined for the MOOC format. In view of the results, which build on existing concepts, it can be stated that this work followed a trend where GT is used as a technique of analysis, rather than a complete methodological approach where the generation of new theory is the ultimate objective (Stough & Lee, 2021).

4. Results and discussion

This section presents the results, which entail defining and outlining the typologies of LMOOCs (concepts developed) and providing accurate definitions that delimit their distinct characteristics based on the data analysed. The presentation of results is accompanied by a discussion of these to allow an overall assessment of the phenomenon under study by justifying the existence of the emerging categories. The application of GT has allowed us to broaden the definition of LMOOCs by conceptualising, as a result, the typologies eligible for categorisation and a comprehensive definition shaping them. Table 2 summarises the main findings, including the final label assigned to each LMOOC category, definitions, number of courses and corresponding percentages.

Table 2. Results: Identified conceptual categories in LMOOCs (n = 430)

It should be noted that the non-inclusion of potentially eligible categories for this taxonomy referred to in the theoretical framework was motivated by different factors: on the one hand, the lack of related evidence in the selected sample and, also applicable to sampling issues, the impossibility of the application of the methodology; on the other, the non-compliance with the basic criteria for the selection of LMOOCs according to what the CEFR understands as language learning.

Such is the case of the LMOOCs using foreign language for instructional purposes, which was excluded from the taxonomy given the absence of sufficient evidence to be raised as a category within it. The analysis of the only two related courses found that could potentially be raised for categorisation did not permit a clear and solid pattern to be traced. However, these courses first emerged with distinctive characteristics in relation to other LMOOCs. They used the foreign language to work on specific subject-matter topics that could both assess language and content or language exclusively. While this work has not included this category, it is susceptible to being further approached in future research with a broader sample of related LMOOCs.

The authors acknowledge that the study has some limitations, which are related mainly to sampling issues. The selection of the sample is reduced to the database of Class Central’s platform, which evidences that the final sample does not correspond to the total number of existing LMOOCs, even if it reflects the courses offered from the principal MOOC providers from reputable institutions around the world.

Another limitation also identified has to do with the platforms’ procedure for keeping courses visible. Class Central takes “… the Bayesian average of ratings … and [removes] closed courses and those with only a small number of reviews” (Shah, 2022). This implies that related evidence already removed from the platform could not be considered. Both situations may have limited the elaboration of categories that have potentially been discussed and proposed as distinctive in this field.

4.1 General language learning LMOOCs

This typology of LMOOCs was the most widespread, accounting for 69.3% of the courses reviewed. Its emerging definition was set as follows: LMOOCs that allow for the development of all communicative language competences on all language proficiency levels in a transversal manner and not subjected to specialised domains. This category has an inclusive nature since it embraces all levels of competence and specific features of each competence, as long as they are applicable to the full set of activities and strategies of communicative competence (e.g. vocabulary, grammar, etc.).

The LMOOCs included in this category had three recurrent common features that allowed us to profile this category in relation to the CEFR (Council of Europe, 2018). First, there was an explicit mention in most of the courses of the competence level worked on, stating its equivalence to the CEFR or, less commonly, other reference frameworks, or using terms that help to locate the level required/worked on (e.g. introductory, intermediate, advanced, etc.) (see Figure 2). The second feature had to do with the transversality mentioned, as courses were aimed at working on all language skills in the foreign language (e.g. listening, speaking, etc.), and had no preestablished potential users. As a result, transversality also functioned as a discriminatory tool in the classification process.

Figure 2. Display of an LMOOC for general language learning. Source: https://www.classcentral.com

It can be objectively stated that certain recurrent characteristics in the courses classified in this modality, such as the information on the level or skills worked, can be found in other emerging categories too. The main difference lies in the fact that the rest of the categories have more exclusive eligibility, whereas the generalist approach of this category includes courses with non-limited conditions, permitting an easier inclusion.

This category can be therefore considered a catch-all for LMOOCs that do not have a predefined target audience, as is the case of LSP LMOOCs; and without recurrent content specifications that merit particular attention, as is the case for culturally or metacognitively based LMOOCs; or the ones that use the foreign language as a medium of instruction, all discussed as follows.

4.2 Language for specific purposes LMOOCs (academic, professional, and one specific skill development)

The definitions of the subcategories emerging from the analysis of the courses with this denomination (21.8%) did not differ from the one that is widely accepted as applicable to other modalities in language learning in the same contexts (Dudley-Evans & St. John, 1998; Johns & Dudley-Evans, 1991; Strevens, 1977).

However, and considering the variable and absolute features of this modality theorised by Johns & Dudley-Evans (1991) and Strevens (1977), the analysed courses were categorised into three sub-strands based on the provision of content related to specific disciplines, activities and occupations, and the specificity provided by the variable feature where just one specific language skill is developed.

Distinctions among LSP-related courses have been historically addressed, which caters for the different contextual linguistic needs represented by absolute variables within this typology (Dudley-Evans & St. John, 1998; Johns & Dudley-Evans, 1991; Strevens, 1977). The novelty of this study was that the variable characteristics of the courses regarded as LSP were also considered to make an additional distinction. This made it possible to differentiate courses that dealt exclusively with one activity or communicative strategy responding to specific learning needs supported by adapted content for its correct development, which may be considered a specific purpose itself. Theory has already considered this differentiation, so the emerging categories were justified based on the research criteria, which made conceptualising the distinctive characteristics of these types of LMOOCs possible.

4.2.1 LMOOCs for academic purposes

This modality comprised 3% of the total sample and 13.8% of the LSP-dedicated courses. Its emerging definition is LMOOCs providing specialised linguistic content used in academic contexts, suitable for all language levels in related disciplines. It can be observed that all the courses were aimed at the university level, and although occasionally descriptions do not explicitly mention this fact, this information was inferable. This category includes courses for the learning of linguistic content, concepts, dynamics, and skills in the foreign language usually used in this context.

These courses also include training on how to communicate effectively in written and spoken interactions, how to understand academic texts in the target language or how to acquire language competences to be able to take and complete graduate-level courses. With a more specific focus, several courses were aimed at learning and developing academic writing with the study of specific patterns in this area. Finally, there were also courses for learning, mainly English language, in specific disciplines such as the ones encompassed in STEM (science, technology, engineering, and mathematics), with this technical area having a strong presence in the sample (see an example in Figure 3).

Figure 3. Display of an LMOOC for academic purposes. Source: https://www.onlinecourses.nptel.ac.in

4.2.2 LMOOCs for professional purposes

These courses are described as LMOOCs providing specialised language content used in professional contexts, appropriate to all language levels in specific work contexts (see Figure 4). It was found that a large majority of LMOOCs categorised for professional purposes (15.5% of the total sample and 71.1% of LSP courses) were specialised in business contexts. Business-related courses were mostly aimed at the development of communicative competences.

Figure 4. Display of an LMOOC for professional purposes. Source: https://classcentral.com

Although these usually approached foreign language learning from a transversal perspective, there were more specialised courses focusing on vocabulary and grammar or written and oral expression, although the most recurrent objective was to achieve effective communication for business. More discretely, there was evidence of courses aimed at other occupations, such as tourism, journalism, or engineering. Finally, it is worth highlighting the recurrence of courses that are prepared for the workplace, such as the preparation of job interviews or CVs.

4.2.3 LMOOCs focused on a specific language skill development

These courses accounted for 3.2% of the total sample and 14.7% of the LSP courses. They are LMOOCs focused on the development of a single language skill or on exclusive features only concerning those (e.g. pronunciation), on condition that they are aimed at specialised domains.

The last part of the description is intended to highlight the fact that this category owes its existence to the absolute characteristics of LSP: the need to meet learners’ needs, the provision of content for specific domains and the adaptation of language at all linguistic levels to the discipline to which it is related.

The courses found in the review embraced all language skills; however, those of production (writing and speaking) were the most common (see description in Figure 5). The exception in this matter was found in a course aimed at developing listening skills that additionally employed the term “audiovisual reception” (watching TV, film and video). Referring to specific language activities and strategies was a novelty found during analysis and it reflects the change of replacing the term “language skills” in the more recent versions of the CEFR.

Figure 5. Display of an LMOOC for the development of one specific language skill. Source: https://classcentral.com

4.3 Cultural-oriented LMOOCs

This typology accounted for 6% of the sample. They are defined as LMOOCs where the cultural elements of the target language are explicitly included in the syllabus and accompany, by all means, the development of communicative competences in the foreign language. The cultural element can be manifested in all its acceptations, considering humanistic and anthropological perspectives, such as sociocultural elements, literature, gastronomy, mythology, philosophy, history, etc.

The classification process for this category considered an evident influence of the cultural element in the content of the linguistic materials (McKay, 2004). Following the model proposed for efficient integration of cultural and linguistic knowledge (Stern et al., 1992), this influence was manifested through the inclusion of elements associated with the two approaches to culture mentioned earlier.

Since the relationship between language and culture is intrinsic in nature (Byram & Wagner, 2018; Choudhury, 2014; Marhamah et al., 2017), it was considered necessary to differentiate between courses that, assuming their communicative approach, worked implicitly on the cultural element and those that decided to approach certain aspects in a more tangible manner. This was reflected in the inclusion of self-evident cultural information in all sections of the courses and leaving aside those who dealt with it occasionally (see example in Figure 6).

Figure 6. Display of the description of a cultural-oriented LMOOC. Source: https://ocw.mit.edu

Cultural explicitness in the configuration of the courses, therefore, determined the eligibility of LMOOCs for this category, as it was understood that the cultural content was the common thread running through the rest of the dynamics during the course (activities, assessments, etc.), as Wang-Szilas and Bellasen (2017) suggested. This way, the orientation of the content is a key element to consider for this category, as it acts as a differentiating feature with respect to other LMOOCs.

4.4 Language meta-learning LMOOCs

Language meta-learning courses represented 2.3% of the sample. A suitable definition of this modality is LMOOCs where foreign language learning relies entirely on linguistic consciousness or prior acquisition of knowledge that enables control and awareness of the language learning. This includes self-awareness of how an individual learns, particularly awareness of learning strategies, and behaviours applicable to the specific learning context (Boström & Lassen, 2006; Jackson, 2004).

The analysed LMOOCs precisely employed these strategies to gain knowledge about the characteristics of the target language (linguistic consciousness). Thus, it is understood that purely linguistic knowledge could be subsequently acquired with increased efficiency. Linguistic consciousness was aimed to be acquired through understanding and assimilation of the rules of the language, learning techniques, strategies, and tricks for the development of language skills, the knowledge of the structural differences between native and target languages, and the resolution of common doubts that allow reflection on the content (see the description in Figure 7).

Figure 7. Display of a language meta-learning LMOOC. Source: https://www.classcentral.com

In this modality, it was observed that very specific aspects of the competences were addressed, mainly linguistic (grammar, pronunciation, etc.). It is important to underline that the courses under this category were dedicated both to the development of metacognition for language learning and to the practice of the target language. Metacognitive knowledge was always accompanied by dynamics that allowed linguistic practice that was conducive to the acquisition and development of communicative competences.

Precisely, the definition of this emerging category was based on this common element, since the definition of theoretical categories needed to be shaped by an exhaustive review of their properties to find the common elements that allow them to be defined (Charmaz, 2012; Glaser & Strauss, 1967). In this way, the eligibility of this category was confirmed, proving the fulfilment of the established criteria, which made it possible to regard them as language courses, as McLoughlin & Magnoni (2017) suggested.

5. Conclusions

The aim of this study was to map the different types of LMOOCs provided nowadays, understand their similarities and differences and, based on the obtained data, establish a solid taxonomy for LMOOCs. After having systematically analysed the corpus retrieved from the brief descriptions and the syllabus of the final sample of 432 LMOOCs, and by applying GT, a taxonomy was created. This taxonomy distinguishes six categories – general language learning LMOOCs, LMOOCs for academic purposes, LMOOCs for professional purposes, LMOOCs focused on a specific language skill development, cultural-oriented LMOOC, and language meta-learning LMOOCs – and describes their respective definitions.

The limitations discussed, however, should be considered for refining this empirical-based taxonomy in future research with the aim of advancing the conceptualisation of the ways to learn foreign languages in MOOCs. It would be desirable, therefore, that LMOOC researchers aiming to contribute to this end take into consideration related research and report their own experiences to contribute valuable evidence that would support the creation of new categories. The extension of the evidence to be tested would need to be supported by rigorous research designs that allow for uniform criteria to be applied to data from different sources.

Despite its limitations, this study means a significant contribution to the LMOOC research field, inasmuch as it is one of the first attempts in the research field to create such a taxonomy and, consequently, to broaden the definition and to build a more robust conceptualisation for LMOOCs through a systematic analysis of content. Moreover, such categorisation has the potential to help LMOOC designers to come up with new, more specific and more efficient design solutions for the different types of MOOCs. Last but not least, the applied research design enriches the MOOC and LMOOC research methodology landscape and might serve as an inspiration to other researchers in the field to apply this particular research design more frequently.