2.1.1 Culture learning

An important strand of research has recognized that blogs are ideal spaces for culture learning and intercultural exchange, but present some challenges. For example, Ducate & Lomicka (Reference Ducate and Lomicka2008) had 9 American university-age learners of French and 20 of German first follow and present on specific bloggers from the target culture and then develop and interact through personal blogs. In their analysis of student work, the researchers found that some students expressed themselves more freely in the blogs than in class or with other assignments. Aligning with future findings in work on learner autonomy (cf. Alm Reference Alm and Thomas2009, §2.1.3) and issues of task design (cf. Lin, Groom & Lin Reference Lin, Groom and Lin2013; Chen Reference Chen2015, §2.1.4), however, Ducate & Lomicka uncovered several challenges, including that students reported not reading each others’ comments, that some learners complained the target blogs they read weren't always topically related to course content, and that some students wanted more freedom in choosing what to write about on their own blogs.

At the same time, Elola & Oskoz (Reference Elola and Oskoz2008) reported no serious challenges in their telecollaboration project where L2 Spanish learners in the US interacted with L2 Spanish learners studying abroad in Spain in small group blogs, with the students abroad acting as cultural informants. They also found little significant differences in the groups before or after the experience with regards to both intercultural competence and attitudes towards blogs as learning tools. In a similar project, Lee (Reference Lee2009) used blogs and podcasts for a telecollaboration project connecting graduate teacher trainees in Spain with Spanish learners in the US. Lee found students liked having an authentic audience, but because of unequal status regarding language proficiency, some students reported a sense of imbalance—an issue not uncommon to telecollaboration. Lee's later project (Reference Lee2012) combined some of Ducate & Lomicka's (Reference Ducate and Lomicka2008) and Elola & Oskoz’ (Reference Elola and Oskoz2008) task designs, as well as ethnography-informed interviews of informants and experts. Lee argued there was evidence the blogs supported development of critical intercultural awareness, particularly because students could see how their own perspectives had changed by comparing old with new entries. Task design may be particularly important when using blogs as learning spaces for learners studying abroad. Gabaudan (Reference Gabaudan2016) notes the need for clear task design, guiding rubrics, and consistent teacher feedback, as well as the importance of scaffolding learners into productive L2 reflective writing practices well before departure.

Moving beyond pedagogical innovation, some later work utilized increasingly sophisticated methods to show evidence for blog-enhanced culture learning, with important caveats. For example, Yang (Reference Yang2011) identified tensions and key events in group blogs created telecollaboratively by 9 English as a foreign language (EFL) learners in Japan and 20 in Taiwan. She found most students interpreted the exchanges from commonality perspectives, believing that mutual understandings and shared interests would lead to rapport, and conversely, that lack of commonality would mean there was no basis for building it. In contrast, students coming from relativity perspectives sought connection by interpreting meanings not at face value but by relating to them, finding commonality on a deeper symbolic level. Yang's work shows that whether the blog-enhanced feature of the exchange afforded learning or not is perhaps beside the point, because blogs (and other social media—cf. §4.1.1) have become normalized as interactional spaces. In other words, while technology may offer new means of interaction, it may mislead because of mediational effects (Ware Reference Ware2005; Kern Reference Kern2014), and transformative learning outcomes are in no way assured.

2.1.2 Literacy and identity development

Another research strand in blog-enhanced L2TL focuses on the development of academic literacy and identity, often using qualitative case study methods in the tradition of Lam's (e.g. Reference Lam2000) work on CMC-mediated identity development. In a highly cited early piece, Bloch (Reference Bloch2007) presents evidence that his focal learner, an adolescent heritage English learner, found space for reflection, self-presentation, and personal expression in the class blog, and thereby developed critical literacy. Similarly, Gebhard, Shin & Seger's (Reference Gebhard, Shin and Seger2011) study shows how a young English learner's use of a class blog for a range of academic and social purposes expanded her communicative repertoire, developed literacy skills, and built meta-linguistic and genre awareness. Also using similar methods and asking similar questions, Sun & Chang (Reference Sun and Chang2012) examined the role of blog-enhanced collaborative dialogue in developing authorial identity among seven Taiwanese teachers of English as second language (TESL) graduate students. Analysis of language use that indexed identity and positioning showed learners were very self-conscious of their inexperience as academic writing novices, questioning their own competence, but finding inspiration in provided narratives by non-native English speaking TESL professionals. Also exploring teacher identity development, Arshavskaya (Reference Arshavskaya2017) found that transformative learning can occur when teacher trainees reflect on triggering episodes and their future identities through blog-enhanced tasks, although some learners may respond better to discussion and more interactive mentorship styles that blogs do not support.

While methodologically rigorous, the facts that Bloch (Reference Bloch2007) and Gebhard et al. (Reference Gebhard, Shin and Seger2011) presented single case studies and that Sun & Chang's (Reference Sun and Chang2012) and Arshavskaya's (Reference Arshavskaya2017) participants were teacher trainees and thus perhaps self-selected do not necessarily make the case for wide-scale application of blog-enhanced L2 pedagogy. Still, they are quite rigorous methodologically and serve to inform practice in relatable situations. Also importantly, even though it was not the focus of the research, a common implication is that successful blog-enhanced L2 instruction may depend on complex interactions among task designs and learner variables (cf. §2.1.5). In other words, the most successful blog tasks encourage reflection, self-presentation, deep knowledge investigation, and development of expertise, but learner variables like background and experience with technology, as well as task variables like audience and topic, may impact outcomes.

2.1.3 Learner autonomy

Another strand of research on blog-enhanced L2TL argues that blog-based tasks should support and develop learner autonomy. For example, Alm (Reference Alm and Thomas2009) designed blog-enhanced German instruction with the goal of enhancing learner autonomy. Her design, informed by self-determination theory (Deci & Ryan 2007, in Alm Reference Alm and Thomas2009: 208), sought to provide optimal challenge and reward, afford social interaction, and promote agency by allowing students to control privacy settings and decide who their audience would be. In a similar piece, Guth (Reference Guth and Thomas2009) described a blog-based ‘personal learning environment’ (PLE) for university EFL learners which could be customized by individual learners, maximizing learner autonomy and serving as a space for collecting, archiving, and managing online learning resources. However, unlike a learning management system, a PLE is open source and designed to be portable and extendable beyond one course, thus promoting informal, autonomous, and lifelong learning.

Later work adds nuance to this understanding with empirical support. For example, Lin, Groom & Lin (Reference Lin, Groom and Lin2013) sought to discover why many students did not blog more than required by a formal curriculum, and did not continue blogging after instruction. Surveys of eight EFL first-year university students showed that they acknowledged the novel, convenient nature of blogging, but that they were more focused on the assigned, formal nature of the learning tasks rather than their blog-enhanced quality. They felt blogs exposed them to new, useful linguistic input, but they also were troubled by perceived social and time pressure to not make linguistic errors in their posts and to offer peer feedback. Students found blogging in class to be a distraction from valuable face-to-face interaction time, but blogging outside of class to be difficult because of other distractions. To explain the results, the researchers hypothesize that their adolescent students’ low proficiency and possibly over-familiarity with social media contributed to dissatisfaction.

2.1.4 Blog-enhanced audience awareness

Another focus of blog-enhanced L2TL research is on development of audience awareness—argued to be an affordance of other social media as well (cf. §3.1.4). In an early piece, Raith (Reference Raith and Thomas2009) compared the paper-based and blog-based journals of 29 adolescent EFL learners for interactive elements and interpersonal language, finding that the blog writers wrote more, and more interactively. A survey then found evidence the bloggers had a more developed sense of audience, while the paper journalers were more concerned with privacy. Arslan & Sahin-Kizli (Reference Arslan and Sahin-Kizil2010) compared blog-enhanced process writing instruction with face-to-face process writing. Comparison showed significant differences in improvement by the experimental group, with the biggest differences in content and organization. The researchers explain this finding in terms of audience awareness—that the bloggers benefited from the knowledge of, and experience with, an expanded audience for their compositions.

However, like the aforementioned nuance that Lin et al.’s (Reference Lin, Groom and Lin2013) work added to understandings about learner autonomy, more recent research also implicates task design in purported affordances for audience awareness. To illustrate, Chen (Reference Chen2015) had 33 Taiwanese, university EFL learners keep personal blogs on topics of their choice, and comment on each others’ blogs. Post-instructional analysis showed that learners as a whole were most satisfied with epistemic (knowledge-focused) activities involving form and grammar correction more than social (knower-focused) activities involving communication and interaction—perhaps because blogs support individual expression rather than collaboration (Warschauer & Grimes Reference Warschauer and Grimes2007). However, further analysis showed that results were potentially skewed because the most satisfied learners exhibited more expertise in the topics of their blogs, and more investment in self-presentation to an imagined audience, even when their posts received few comments (cf. King Reference King2015, §3.1.4). Unsatisfied learners felt less invested in, or less qualified to write about, their blogs’ topics, and were either negatively impacted by audience awareness, like some of Lee's (Reference Lee2009, §2.1.1) students, or convinced there was no audience for their blogs, because they were mere class assignments. In short, tasks that emphasize an external audience seem to be a double-edged sword, because awareness of that audience can both empower and intimidate L2 learners.

2.1.5 Blog-enhanced task design and learner variables

Research on blog-enhanced intercultural exchange, literacy and identity development, learner autonomy, and audience awareness implicates consideration of both task design and learner variables in realizing the learning affordances of blogs. Dippold's (Reference Dippold2009) study of peer feedback in blogs illustrates how the two are intertwined. Examining the quality of blog comments advanced German learners provided each other, the researcher found they were rarely critical or focused on form as learners were hesitant because of face issues and insecurities about language abilities. In an advanced academic EFL course with 11 students, Vurdien (Reference Vurdien2013) found evidence that well-designed blog-enhanced writing tasks complemented by face-to-face discussion led to self-assessed learning gains. While most students appreciated the compositional and reflection time afforded by asynchronous pre-writing and feedback tasks, a few disliked their structured, imposed nature, and felt more open-ended discussion activities related to the current issues they wrote about would have been more motivating. Implications from Dippold and Vurdien's studies are that learners should have some choice in the topical focus of social media activities, in addition to some control over audience (cf. Chen Reference Chen2015, §2.1.4) to alleviate possible face issues.

Chen, Shih & Liu (Reference Chen, Shih and Liu2015) focused directly on blog-enhanced L2 learning task design, which they argue may be responsible for some negative findings, like those of Krause (Reference Krause2005), who found that vague task parameters demotivated students, blogs were often used where other simpler technologies would do, and blog-enhanced peer feedback was ineffectual. Working with 17 student dyads in an intra-national Taiwanese EFL telecollaboration partnership, they compared open and closed dialogue tasks—the former does not require negotiation or decision making, while the latter requires a single joint product as an outcome. Closed tasks elicited more frequent generation of idea units and were preferred by students, even though they felt the blog was a better medium for open tasks that supported divergent thought and depth of opinion. In other words, learners themselves recognize when blog-based tasks do not align with blog affordances, even if the tasks align with their own learning styles and preferences.

2.1.6 Summary and future directions: Blog-enhanced L2TL

In short, findings from early studies on the affordances of blog-enhanced L2TL have become more nuanced over time with caveats from later studies, especially regarding task design and the influence of learner variables. Research has shown some evidence that blogs may serve as spaces for culture learning and intercultural exchange, as well as for reflection, self-presentation, and the development and expression of deep knowledge and expertise on topics in which the learner has had some say. Later research has found that tasks which force particular topics, require form-focused peer review, restrict audience, broaden audience, or are closed in nature may restrict learner autonomy and not support the learning potentials of blogs, even though those tasks might more align with curricular objectives and learner expectations.

In the future blogs may continue to serve as learning spaces for writing oneself into multilingual and intercultural identities. In addition, there are research lacunae on multimodal composition and digital storytelling, as blogging overlaps and integrates with image sharing social media like Instagram, Pinterest, and Tumblr. As new areas are explored, it should be noted that in general, because of familiarity with everyday social media, learners may have experiences and dispositions that should be taken into consideration when designing and implementing instruction. On the other hand, as blogging becomes a niche practice, like wikis, it may require more explicit instruction, at the same time its novelty offers advantages as a sort of tabula rasa.

3.1.1 Meaning over form

The first strand of research has been on whether wikis are appropriate for focus on accuracy of form, since in non-educational, informal contexts they function as tools for collaborative knowledge building and communication where focus on meaning is paramount. In one of the first studies to ask the question, Kessler (Reference Kessler2009) reports on a project involving a wiki created by 40 L2 English teacher trainees, without any instructor involvement, that was supplemental to a course on culture. Analysis of the wiki content and version history showed that the students focused far less on accuracy of form than on meaning in peer corrections, which Kessler reasons was due to the informal nature of the task and the wiki environment, and to the inconsequential nature of the errors for effective communication. Kessler's results have been cited frequently as evidence that without intervention, wikis do not promote focus on form.

In a study of the wiki revising activity of three classes of 54 L2 German learners, Arnold, Ducate & Kost (Reference Arnold, Ducate, Kost, Lomicka and Lord2009) found that the learners made more meaning-focused changes than accuracy-focused changes overall, while making formal revisions to their own work and leaving others’ work untouched. While the learners corrected about three-quarters of their own errors autonomously and revised their wikis much more frequently than studies of paper-based revising had found, the two classes that received more teacher guidance produced more successful revisions than the other section. In other words, students may not offer revisions to peers unless instructed to do so, especially when they perceive their teacher is the true audience—a finding echoed in research on other social media (cf. Dippold Reference Dippold2009, §2.1.5; Chen Reference Chen2015, §2.1.4). While Arnold et al.’s (Reference Arnold, Ducate, Kost, Lomicka and Lord2009) work complements Kessler's, it is sometimes cited as evidence that learners can, and do, focus on form in wikis, although the pivotal role of teacher guidance is sometimes glossed over.

Finally, in a study reporting on the same intervention as Kessler's aforementioned earlier work (Reference Kessler2009), Kessler & Bikowski (Reference Kessler and Bikowski2010) examined the kinds of meaning-focused changes their participants made in their wiki. They found that only five of the 40 students made more than three changes, with about half making only one, and that most changes involved creation, deletion, or elaboration rather than synthesis. The researchers conclude that while there was some indication of what they call ‘collaborative autonomous learning’, many students used the wiki only intermittently, making edits that required little critical thinking, and that most used it only because it was a requirement. Echoing other findings with wikis (cf. Bradley, Lindstrom & Rystedt Reference Bradley, Lindstrom and Rystedt2010, §3.1.2) and other social media (e.g. Vurdien Reference Vurdien2013, §2.1.4), they suggest that for wikis to support development of learner autonomy, activities should be flexible and student-directed, even though participation may be uneven and outcomes unpredictable.

3.1.2 Collective activity, collaboration, and cooperation

Since a major purported affordance of wiki-enhanced L2TL is for collaborative writing (Storch Reference Storch2011), a second point of investigation has operationalized and contrasted the concepts of collective activity, collaboration, and cooperation. In one of the first widely cited studies on the topic, Lund (Reference Lund2008) offered a sociocultural account of how collective activity emerges as 30 English learners in Norway planned and created a class wiki. In his analysis of two dyadic interactions, Lund described how the unfolding activity shifted from localized, individual ownership to collective, interdependent production reflecting values of reciprocity, multi-voicedness, and trust. Lund clearly situates his study in sociocultural theory and methodology, being careful not to conflate the terms ‘collective’ and ‘collaboration’, and in fact never uses either ‘collaboration’ or ‘cooperation’—a nuance not always recognized by subsequent researchers citing him as evidence for the purported collaborative benefits of wikis.

Bradley et al. (Reference Bradley, Lindstrom and Rystedt2010) note critically that ‘the potential of the wiki concept as a writing tool is frequently assumed’ (p. 249). They had 25 dyads and triads of university-level English learners in Sweden complete wiki-based writing tasks designed to practice differing genres and registers. Analysis of the wikis shows varying degrees of cooperation and collaboration, with 10 groups showing only cooperation or no interaction at all, and 15 showing collaboration as evidenced by true joint authoring and mutual feedback on both form and meaning. The authors imply that collaboration can be encouraged through situated learning task design but that for students to remain motivated and retain agency, variance in participation levels should be anticipated and permitted—an important caveat that mirrors similar findings on blogs.

Arnold, Ducate & Kost (Reference Arnold, Ducate and Kost2012) examined their aforementioned 2009 data for examples of true collaboration as opposed to cooperation, defining the former as unified attention by all participants on the same work, and the latter as separate work in unified spaces. In their new analysis, the researchers found evidence for four assumed roles: free rider, social loafer, team player, and leader. They found these assumed roles may lead to uneven distribution of work and lack of collaboration, sometimes precipitated by the ‘first responder’ phenomenon where the work of the author first to contribute to a wiki is least likely to be altered.

True collaboration may also depend on willingness to share tasks and shift roles between tasks. In a study of 90 university EFL learners in China, Lai, Lei & Liu (Reference Lai, Lei and Liu2016) found groups who chose to share both wiki composing and co-revising tasks—rather than some composing and some revising—were more satisfied overall with their experiences than those who did the tasks individually. Li & Zhu (Reference Li and Zhu2017) examined the interactions of two groups of three ESL learners through and around two wiki-based collaborative assignments for evidence of equality and mutuality, finding variance in the groups as individual motivations, roles, and emotional stances shifted between tasks. Pedagogical implications are to have groups self-assess their collaborative efforts in terms of these aspects throughout the process.

3.1.3 VLEs

Because wikis are easily customizable, they can be used as VLEs that provide supplemental resources and scaffolding on how and why to use wikis for L2 learning. For example, deHaan et al. (Reference DeHaan, Johnson, Yoshimura and Kondo2012) had 13 university-level English learners in Japan use a wiki to prepare, transcribe, discuss, and practice spoken role plays, focusing on interactional competence and fluency. The researchers report overall improvements in grammar, pragmatics, vocabulary, and learner confidence. Even though tasks were highly structured and the wiki did not have an external audience, the project was perhaps successful because students saw themselves as the wiki's audience and recognized that the task parameters aligned with what the wiki could do (cf. Pellet Reference Pellet2012, §3.1.4).

However, projects may fail if the tasks are not perceived as aligned with learner needs and wiki affordances. For example, while most wiki projects are implemented at the intermediate level or above, Kennedy & Miceli (Reference Kennedy and Miceli2013) sought to use a wiki for their university-level beginning Italian learners at three Australian campuses. The wiki served as a common space for notice boards, discussion, social networking, and resource sharing. A post-instructional survey showed that only a minority of the 79 students enjoyed the wiki and felt it was beneficial to their learning. Qualitative responses showed appreciation for interaction, but frustration at technical problems, lack of training, wiki etiquette, division of labor, and lack of teacher presence. The researchers implicate the inclusion in any wiki-based VLE of resources that focus on reflective practice and critical awareness about collaboration in language learning, especially when beginning level learners are involved.

Recounting similar projects, Rott & Weber (Reference Rott and Weber2013) offer suggestions based on their experiences using wikis in L2 German and French instruction. They explain that guided discussion and exploration of an example wiki along with templates can give students a better understanding of wiki structures, while requiring reading of each others’ contributions can help build awareness of audience. They also maintain that students benefit from discussion of different genre types found in wikis—descriptive, interpretive, and summative writing—as well as how to use wiki software, how to exploit in-wiki communication tools, how to take and share notes, and how, when, and when not to use translation tools. In short, their piece offers guidelines for situated and awareness-focused instruction—approaches found effective in the development of literacies in other social media as well (cf. §4.2.3 and §4.2.4).

3.1.4 Wiki-enhanced audience awareness

Like findings regarding the potential of blogs to develop audience awareness (§2.1.4), findings on audience awareness in wikis come with caveats. An early piece on awareness in wikis that is cited widely, although not always critically, is Mak & Coniam (Reference Mak and Coniam2008), who had Hong Kong high school EFL learners use a wiki to develop a school brochure for their parents. The researchers found that over the course of the project, focal students on average produced far more language output in the wiki than the minimum required, and that the output was longer in terms of t-units. They explain the improvement in terms of wiki affordances for collaboration and processes of revising, expanding, reorganizing, and correcting, with the motivation of a potentially real audience of readers. The results should be taken judiciously, however, since Mak & Coniam presented results only from 4 of 24 students, chosen because their wiki was voted best by teachers (p. 444); it is unknown how the project went with the majority of the students.

There may be potential when students understand themselves as their own wiki audience (cf. deHaan et al. Reference DeHaan, Johnson, Yoshimura and Kondo2012, §3.1.3). For example, Pellet (Reference Pellet2012) had 30 advanced French sociolinguistics undergraduate students create wikis to build, manage, and share course content knowledge—in French, directed by the students, without focus on grammatical accuracy (perhaps taking Kessler Reference Kessler2009, §3.1.1 into account). Students responded very positively to the project, and recognized the wiki's role as a collaborative, agency-rich space for critical discussion and knowledge building. It is worth noting that this purpose—using a wiki as a shared knowledge resource or repository like a dictionary, glossary, guidebook, or encyclopedia—is also the most common outside of educational contexts. It should also be noted that as advanced French majors at a small liberal arts college, Pellet's students were also self-selected and highly motivated.

Incorporating this authentic purpose may account also for the successes of King's project (Reference King2015), where university-level Hong Kong English learners collaboratively contributed Wikipedia entries on various Hong Kong subjects and reflected on how the experience impacted their identities as English users. Some students reported feeling a sense of writer responsibility, obligation to a wider Wikipedia community, and satisfaction when their entries were not deleted, while others demonstrated a sense of legitimate participation when discussing with other Wikipedians the status of their contributions. King argues that even just the potential of a genuine public audience afforded users the development of authorial identities in imagined communities, and while Wikipedia has since replaced or deleted all of the student entries, the experience for some may have been transformative. As King only presented selected case studies, however, it remains unknown whether any students felt intimidated, rather than empowered, by a potential external audience, as has been found is possible with blogs (cf. Chen Reference Chen2015, §2.1.4).

3.1.5 Wiki-enhanced task design and learner variables

As research on focus on form and collaboration in blogs has found (§2.1.5), task design and learner variables also play a major part in realization of wiki learning affordances. Elola & Oskoz (Reference Elola and Oskoz2010) had eight advanced Spanish learners use wikis and chat to complete both individual and group assignments, and compared outcomes from different conditions. Differences emerged in how individuals and groups structured and organized essays, and at what stages in the writing process they made local edits. While students recognized the value that collaborative writing afforded in terms of learning structure and organization, they felt the wiki was less useful for learning grammar than initially expected—not particularly surprising considering other findings (cf. Kessler Reference Kessler2009, §3.1.1), but notable in that students’ own expectations were not met, perhaps because in this case they did not have previous wiki experience.

In a similar study, Lee & Wang (Reference Lee and Wang2013) analyzed the wiki project participation of 103 English learners at two different Taiwanese universities over 18 weeks to determine what factors facilitated and hindered participation, including the nature of the task, student working style preference and orientation towards peer review, and instructor attitudes towards technology. Quantitative and qualitative analysis of student work, surveys, and interviews showed findings similar to Arnold et al. (Reference Arnold, Ducate and Kost2012, §3.1.2), where successful groups exhibited even workload distribution, mutual respect of opinion, and sustained interaction. Hindrances to success included the opposite—uneven distribution, lack of mutual respect, and unsustained interaction—and were partially attributed to technological and contextual constraints like time delays.

As would be expected, task type also plays a role in the type and amount of collaboration. Building on the aforementioned work of Kessler, Arnold, and others (§3.1.1), Aydin & Yildiz (Reference Aydin and Yildiz2014) compared the meaning and form-related changes made by 34 Turkish university-level intermediate English learners in argumentative, informative, and decision-making wiki-based tasks. More changes of both types were found in argumentative tasks, while there were fewest form-related changes in informative tasks, and nearly equal numbers of both in decision-making tasks. Most peer corrections were found in argumentative tasks, while most self corrections were in the informative ones. In their interpretation, the researchers conclude that argumentative tasks were more conducive to collaboration and correction because students felt more compelled to improve opinions with which they agreed, while collaborative negotiation was unnecessary in informative tasks because their outcomes were factual and not contentious. They also observed, although over 90% of the corrections were accurate, that there were very few changes involving synthesis, the hallmark of true collaboration (cf. Kessler & Bikowski Reference Kessler and Bikowski2010, §3.1.1). Moreover, they found half of the students didn't enjoy the tasks—perhaps like Chen et al.’s (Reference Chen, Shih and Liu2015, §2.1.4) blogging students they would have preferred closed tasks, or like Chen's (Reference Chen2015, §2.1.5) students, only information tasks.

3.1.6 Summary and future directions: Wiki-enhanced L2TL

In sum, research on wiki-enhanced L2TL has shown that wikis do not necessarily promote focus on formal revision and the development of accuracy, especially without instructional intervention. As group learning tools, wikis may afford cooperation and shared authorship, but not necessarily true collaboration, and variability in participation and role assumption should be anticipated. Wikis may serve as VLEs, as long as needs and tasks are truly enhanced by wiki structures and awareness of how and why to use wikis is addressed through situated and explicit instruction. Wiki-enhanced instruction may help develop audience awareness if learners believe their collective expertise is worth sharing—with each other or with broader audiences. As with blogs, formal wiki tasks may be more effective if they are authentic and are similar to naturalistic, informal wiki authoring activities, although some learners may prefer traditional activities that align with their expectations of formal learning.

Future research on wikis should continue to examine the potential affordances wikis offer for collaborative authorship and the influence of different task parameters and learner variables on outcomes. Mixed method research techniques that combine analysis of wiki changes (e.g. Yim & Warschauer Reference Yim and Warschauer2017) alongside consideration of learner motivation and stance promise to better capture the dynamic nature of collaboration (e.g. Alghasab & Handley Reference Alghasab and Handley2017; Li & Zhu Reference Li and Zhu2017). Most L2 learners will have very little, if any, experience creating wikis, unlike SNS profiles, and will probably have used Wikipedia much less than they have used SNSs. For that reason, explicit scaffolding and instruction is warranted with wikis, including awareness building regarding their utility as learning and collaboration tools. Building awareness might involve exploring with students the concepts of collective intelligence and crowdsourced wisdom, for example in social media like Reddit or Digg—two popular tools whose application to L2TL has yet to be explored fully.

4.1.1 Strategic maintenance and learning of heritage languages

Researchers have argued that multilingual users and L2 learners find in SNSs tools effective and strategic means for learning and using languages, practicing heritage identities, and maintaining home connections. In one of the first studies along these lines, Lee (Reference Lee2006) presented two case studies of the electronic literacy practices of two adolescent female heritage speakers of Korean in the US who used the Korean SNS Cyworld to practice their language informally through interaction with friends. Research on heritage learners’ use of SNSs has also included bilingual English-Spanish speakers in Puerto Rico (Carroll Reference Carroll2008), and Welsh users in the UK (e.g. Cunliffe, Morris & Prys Reference Cunliffe, Morris and Prys2013; Jones Reference Jones2015). In general, findings are that users strategically consider addressee and purpose in choice of language when code mixing or shifting, and they find community in SNSs among geographically dispersed populations.

The common implication that L2 learners and users find empowerment and voice in informal but strategic SNS use is echoed in other studies that focus more intently on identity and literacy development. For example, in case studies of three college L2 writers, DePew & Miller-Cochran (Reference DePew, Miller-Cochran, Cox, Jordan, Ortmeier-Hooper and Schwartz2010) show how each participant used different SNSs for different purposes, demonstrating varying degrees of symbolic and social networking literacies, including audience awareness. Using similar methods, DePew (Reference DePew2011) offers three more case studies of the extramural Facebook practices of lower proficiency L2 writers, who demonstrated facility with different registers and audiences and mixing languages to present heritage identities. DePew argues that even though L2 writers may be construed as deficient in English proficiency, their highly sophisticated and strategic use of social media is seldom recognized by their schools as a legitimate sort of literacy. Pedagogical implications are to acknowledge these new literacies and to leverage them by somehow integrating SNSs into formal instruction (cf. §4.2.4).

4.1.2 Self-directed learning of foreign languages

Building on the notion that vernacular SNSs are used strategically, another strand focuses on the self-directed, or deliberate and intentional, foreign language learning that they mediate. For example, Pasfield-Neofitou (Reference Pasfield-Neofitou2011) analyzed the online SNS, blog, and email-mediated writings of 12 Australian learners of Japanese and their contacts over 4 years. She examined how language and topic choice indexed identity presentation, conceptions of nationality, and the perceived ownership of online spaces. Participants used their Japanese non-native user/learner and English speaker identities strategically according to interactional purpose, and found some SNSs more accommodating to their learner identities than others. Back (Reference Back2013) analyzed the daily Facebook posts of three learners of Portuguese studying abroad in Brazil and found considerable differences in number, length, and change in length of posts in Portuguese over time. Interpretation implicates the development of audience and register awareness afforded by SNS use, albeit via different pathways, at differing rates, and to different degrees for individual learners. Jin (Reference Jin2018) showed how two L2 Chinese learners studying abroad in Shanghai utilized the ecological affordances of WeChat in different beneficial ways. For example, the lower proficiency learner leveraged the tool's multimodal resources like stickers and memes to scaffold his L2 use, while the higher proficiency learner accessed vernacular registers by virtue of being part of a network of other learners, L1 speakers, and instructors. A common implication of research in this strand is to advocate the use of vernacular SNSs extramurally and to teach explicitly the skills needed for their strategic uses as L2 learners (cf. Prichard Reference Prichard2013, §4.2.3). As surveys (e.g. Lai, Hu & Lyu Reference Lai, Hu and Lyu2018) have found that some learners find social media for informal, extramural learning forced and inauthentic, however, participation should be encouraged in interest-focused affinity spaces where L2 use is not the goal but the means of interaction.

4.1.3 Theoretical and methodological innovation

Descriptive research of informal, multilingual SNS use usually employs qualitative, ethnographic methods in the tradition of Lam's online language socialization research (e.g. Reference Lam2000, Reference Lam2004). Recent work also borrows concepts from research in Internet sociolinguistics that recognizes the multilingual and superdiverse nature of online language use (e.g. Vertovec Reference Vertovec2007; Blommaert Reference Blommaert2010; Androutsopoulos Reference Androutsopoulos2015). For example, Schreiber (Reference Schreiber2015) examined the literacy practices of a Serbian adolescent in Facebook, and argues they should be considered ‘translanguaging’, wherein the user considers the various repertoires and resources as an integrated system, and the SNS serves as an intertextual space for self-presentation and identity performance (cf. §4.1.1). Schreiber showed how her participant was able to develop and present both local and transnational, global English-mediated identities through a variety of translingual re-entextualization and sharing practices that affiliated him with global hip-hop communities.

Other studies employ mixed and quantitative techniques as well. For example, Chen (Reference Chen2013; see also Chen Reference Chen2012) examined the Facebook use over two years of two L1 Chinese graduate student multilingual writers in the US, finding one student preferred to post personal updates and observations while the other shifted from personal updates to sharing and other-involved writing as she observed others doing the same. Working with the second participant's data further, Reinhardt & Chen (Reference Reinhardt, Chen, Lamy and Zourou2013) used frequency analysis to study the participant's development and made the argument that her Facebook activity indexed perceived socialization into an imagined community (Anderson 1991, in Reinhardt & Chen Reference Reinhardt, Chen, Lamy and Zourou2013). Also employing mixed methods, Solmaz (Reference Solmaz2015) used an innovative visualized social network analysis technique to present evidence that students used diverse multilingual and multimodal resources in online identity performance and SNS literacy practices. Visualization showed that an individual's networks mapped into different patterns according to language, gender, and community affiliation.

In a recent book-length treatment, Vandergriff (Reference Vandergriff2016) provides a useful model for applying computer-mediated discourse analytic techniques to a variety of SNS contexts of L2 use and learning in both vernacular SNSs and SNECSs (§4.3). In the examples, she shows how users employ agency through strategic self-presentation, deployment of multilingual resources, negotiation for supportive space, engagement in play, and management of learning processes.

Innovative approaches have led to important classroom implications. For example, Warner & Chen (Reference Warner and Chen2017) adapt Goffmanesque participation analytic techniques to three translingual Facebook conversations, showing that features like virtual adjacency, non-verbal backchanneling, focused addressivity, and layered simultaneity make them highly unique in comparison to non-digital conversations. As such, the authors argue, they might serve better to develop symbolic competence and awareness of symbolic power than the artificial examples found in classroom texts. In their analysis of the translingual practices of their first language (L1) Arabic and L1 Spanish students in an English telecollaboration project, Kulavuz-Onal & Vásquez (Reference Kulavuz-Onal and Vásquez2018) showed that all three languages became interactional resources for audience design, intercultural awareness, and establishing solidarity, arguing that ‘English only’ classroom policies do reflect the global reality of social media literacy practices.

4.1.4 Summary and future directions: Informal SNS L2 use and learning

Research on informal use of SNSs for L2 use and learning differs from research on pedagogical applications of blogs and wikis because there are far fewer, if any, notable studies on the informal use of those tools for L2 use and learning. SNSs are unique in this regard because they are vernacular and everyday, but they are truly global and used in multilingual and transcultural ways. Their use centers on networking with new and old friends, family, and home or heritage culture communities. Unfortunately, the considerable audience and register awareness, interactional, and identity performance skills involved in SNS use are usually neither recognized nor legitimized in traditional writing or language instruction.

Research in the area holds great potential to inform SNS-enhanced L2 pedagogy and to benefit from interdisciplinary cross-pollination. Internet sociolinguistics and computer-mediated discourse analysis are growing fields (e.g. Page et al. Reference Page, Barton, Unger and Zappovigna2014; Jones, Chik & Hafner Reference Jones, Chik and Hafner2015; Vandergriff Reference Vandergriff2016), and offer examples of analyses of computer-mediated and social media discourse with frameworks that can be adapted to L2 situations as well. The field will also benefit from the addition of quantitative techniques that complement the many case studies and ethnographic perspectives, allowing for more generalizable and scalable results. Because SNS use has become so ubiquitous and is digitized, applying mixed methods should not be as difficult as it perhaps once was, although new issues of gaining permission, maintaining privacy, and archiving digital data present additional challenges.

4.2.1 Motivational benefits of SNSs

Early work on vernacular SNS use included idea pieces focused on its motivational potential. For example, McCarty (Reference McCarty and Thomas2009) described his attempt to explore learner motivations and reactions by joining his EFL students’ social network Mixi, where he found mixed reactions from students at his presence. His finding pointed to an early point of debate—whether teachers should be SNS friends with students or force students to friend each other (see also McBride Reference McBride, Lomicka and Lord2009)—that has now largely been solved since Facebook introduced its groups feature. McCarty's study also illustrated another increasingly relevant issue—whether vernacular, everyday technologies like SNSs can be fully appropriated for formal educational purposes and retain their motivational qualities. In an empirical answer to the question, Kelley (Reference Kelley2010) found that four experimental classes of Chinese university EFL learners who participated in MySpace discussions measured significantly higher in some measures of integrative/ideal L2 self-motivation post-treatment than four control classes who did comparable activities face-to-face. While there might have been a novelty effect and the teacher was not controlled for, a plausible interpretation was that the SNS (and the teacher) functioned as a ‘window’ to the outside world, which for Chinese students at the time, was a significant motivator. It would be interesting to test SNS vs. non-SNS conditions experimentally now, several years later, when SNS use has truly become ubiquitous and everyday, even in once-isolated China. While the reason ‘because students are motivated by technology’ might once have been sufficient to justify a technology-enhanced activity over a traditional one, learners may now find classroom technology normalized (Bax Reference Bax2011).

Some recent studies have compared social media-enhanced activities to controlled versions of the same activities face-to-face. For example, Dizon (2016) found that the writing fluency of a Facebook experimental group improved in comparison to a face-to-face control group, but lexical richness and grammatical accuracy did not. Özdemir (Reference Özdemir2017; cf. Jin Reference Jin2015) found improvement in measures of intercultural competence in Facebook discussions compared to control groups. While she did not compare them to face-to-face equivalents, Liu (Reference Liu2017) found more instances of negotiation and resolution in Facebook chats among high-level proficiency learners than low-level, who relied more on requests for help. Although these results are interesting, future research should focus on which particular social media dynamics lead to L2 learning affordances by isolating task and contextual variables. For example, the anytime accessibility, informality, reduced social cues, or anonymity of social media posting may be more associated with some learning affordances than others.

4.2.2 Genuine L2 sociopragmatic usage

One of the first to propose potential formal applications of SNSs, McBride (Reference McBride, Lomicka and Lord2009) describes them as spaces for learner self-authorship, interaction, and sociopragmatic development. Echoing Prensky's (Reference Prensky2001) digital natives argument, she argues that SNS users practice ‘writing/re-mixing of the self’ through impression management, identity work, and identity play, which may all contribute to L2 development. Heralding future studies, she suggests a variety of projects included alternative identity profiles, group profiles, global simulations, media-centered projects, and theme-centered projects.

One strand of research to echo McBride's (Reference McBride, Lomicka and Lord2009) suggestions was forged by Blattner and colleagues, who recognized the potential of SNSs to serve as windows onto genuine L2 sociopragmatic usage. For example, Blattner & Fiori (Reference Blattner and Fiori2009) had university-level French and Spanish learners in the US observe expert speakers informally through Facebook's group function. The instruction, based in Lave & Wenger's (Reference Lave and Wenger1991) notion of situated learning, had learners observe and gather information on how speech acts were used and how usage differed among varieties. In a later study, Blattner & Fiori (Reference Blattner and Fiori2011) had advanced Spanish learners participate in guided discourse analysis tasks designed to raise sociopragmatic awareness of various speech acts and discourse functions in Facebook groups whose subjects corresponded to course topics. Students in both studies responded positively. Finally, Blattner & Lomicka (Reference Blattner and Lomicka2012) reported on the use of Facebook groups as a space for telecollaborative exchange between French learners in the US and English learners in France. Comparison of pre- and post-instruction attitudes towards formal SNS-mediated learning showed a shift towards awareness of the L2 learning potential of the medium. Although they were not required to, many students friended each other and thus remained connected after the course finished.

Twitter has also been explored as a means for exposing learners to genuine usage outside of traditional class times and spaces. For example, Antenos-Conforti (Reference Antenos-Conforti, Lomicka and Lord2009) had her intermediate L2 Italian learners follow each other on Twitter, inviting a few native speakers to participate as well. While most learners reported satisfaction at being able to interact in the L2 authentically outside of class, half reported being overwhelmed with the frequency of the contributions from the native speakers, and discomfort at not knowing them in real life. Lomicka & Lord (Reference Lomicka and Lord2011) had 13 L2 French learners use Twitter to interact with each other and with native French speakers. Analysis found affective (e.g. humor and self-disclosure) and interactive features in the tweets, but fewer cohesive features in the form of vocatives and phatic language, perhaps due to the implied addressivity of a closed community. The invited native speakers did not tweet nearly as much as the learners, which might indicate they did not feel a part of the group. For beginning learners especially, there might be less intimidation if Twitter use in the L2 mirrors the authentic practice of following celebrities, and using the tweets as objects of analysis (e.g. Blattner, Dalola & Lomicka Reference Blattner, Dalola, Lomicka, Dixon and Thomas2015), rather than as a tool for closed group interaction.

Findings on SNS are similar to those of other studies of social media-enhanced cultural exchange, where blogs served as windows onto genuine uses and means of interaction with cultural informants and experts. While not always reported, challenges related to appropriate task design and learner variables—the double-edged sword of peer and authentic audiences empowering some learners but intimidating others—are considerable (cf. Lee Reference Lee2009, §2.1.1; Chen Reference Chen2015, §2.1.4). Studies also sometimes fail to discuss if and how connections initiated in SNS-enhanced intercultural exchange might lead to sustainable autonomous learning practices. Future work might examine this potential, informed by research on how learners use SNS strategically for informal learning outside of the classroom (§4.1, §4.3).

4.2.3 Situated and simulated practice

Also implementing some of McBride's (Reference McBride, Lomicka and Lord2009) suggestions is a strand of research building on the potential of SNSs for situated learning (Lave & Wenger Reference Lave and Wenger1991), since they have become ubiquitous and familiar and might therefore act as a point of departure for exploration of new practices in the L2. For example, Mills (Reference Mills2011) developed Facebook-enhanced global simulation instruction for university-level advanced French learners. Over the course of a semester, students developed simulated identities by completing joint narrative authoring tasks, interacting as their characters in wall posts and status updates, and sharing various online resources, which contributed to a sense of immersion. In the ensuant SNS-mediated interactions and simulated self-presentations, Mills found evidence of the joint enterprise, mutual engagement, and shared repertoire characteristic of situated learning. In a similar project, Reinhardt & Ryu (Reference Reinhardt and Ryu2013) created and implemented a series of role play and bridging activities for intermediate L2 Korean learners, the purpose of which was to develop awareness of Korean Internet language and sociogrammar by voicing perspectives they could not with their actual identities. Learners analyzed each others’ production and discussed the appropriateness of the language used. In another example, Yen, Hou & Chang (Reference Yen, Hou and Chang2015) developed a series of team-based activities for 42 EFL learners in Taiwan enhanced by Skype and Facebook. Teams used Facebook groups as a medium for brainstorming and discussing sales strategies they would then employ in the Skype-based business negotiation role plays. These pieces offer models for instruction by means of which learners can identify with and as L2 users through familiar SNS lenses. In other words, to continue the metaphor of SNSs as windows onto socio-pragmatically genuine discourses and gateways into communities, SNSs can also as playgrounds or safe spaces for practice, perhaps before they are used for interaction with expert users and genuine communities.

4.2.4 Literacies development

As social media becomes increasingly used as a primary source of news and information, there is also justification for focus on media literacy in L2TL (e.g. Reinhardt & Thorne Reference Reinhardt, Thorne, Arnold and Ducate2011; cf. also §3.1.3). These might be integrated with ‘SNS literacies’, which Solmaz (Reference Solmaz2015) argued, based on his descriptive analysis of the informal SNS practices of multilingual university students (§4.1.3), include simulation, performance, appropriation, judgement, and networking literacies. Research shows these activities should be experiential and focused on developing criticality. For example, Prichard (Reference Prichard2013) developed and implemented a set of SNS-enhanced activities corresponding to the TESOL technology standards for 41 Japanese EFL learners. Activities had students experience Facebook in English by creating profiles, friending others, joining groups, and making posts and comments. Similar to Reinhardt & Ryu (Reference Reinhardt and Ryu2013) but using real rather than role played identities, critical activities had students determine the appropriacy and relevance of different posts, and critically evaluate SNSs as aids to L2 learning. Students responded positively to the intervention, although for the most part they used the site more for informal socializing than for explicit study purposes, a finding that aligns with studies of other social media use (Kessler Reference Kessler2009, §3.1.1)

Because SNSs are associated with informal, everyday uses, formal SNS activities may not always be well-received, as Reinhardt & Zander (Reference Reinhardt and Zander2011) found. They had their students play and evaluate Facebook-based social games together and critically evaluate the use of SNSs for educational purposes. Because of the extrinsic pressures of an impending standardized test, however, some learners resisted the SNS-enhanced activities and did not wish to interact with classmates on SNS outside of class, preferring more rote and traditional test preparation activities. Wang & Kim (Reference Wang and Kim2014) found that their advanced Chinese learners for the most part enjoyed an assignment requiring open-ended posts and responses to the group Facebook page, valuing it for extramural practice, freedom of topic choice, and socializing with other class members, but some expressed concern that it distracted them from other more academic-focused assignments. This aligns with findings (cf. §2.1.5 and §3.1.5) that learners themselves may have unexamined preferences for traditional, transmission-oriented, and teacher-driven language instruction. Williams, Abraham & Bostelmann (Reference Williams, Abraham, Bostelmann, Guikema and Williams2014) found half of the 800 American university students they surveyed did not think social media should be integrated into L2 curricula—future research might explore reasons for this and what implications there are for formal uses.

4.2.5 Autonomous learning

Findings on learner pushback on formal, explicit SNS-enhanced instruction have led to exploration of whether there is the possibility of encouraging learners to teach themselves. Studies of semi-formal (sometimes confusingly called non-formal) or extramural uses of SNSs support the hypothesis that giving learners a sense of control is key to developing autonomy as these connections often result in more naturalistic interactions and informal learning opportunities. For example, Liaw & English (Reference Liaw, English, Lamy and Zourou2013) compared the exchanges made between English learners in France and Taiwan in a formal, institution-supported SNS and a Facebook group that they spontaneously created on their own. A systemic-functional linguistic analysis found less interpersonal and textual language in the official group than in the unofficial group. The researchers suggest that the unsanctioned quality of the informal group afforded the learning of interpersonal language not possible in the formal group, leading to more genuine and situated learning of culture.

Another way to encourage autonomy might be to carefully scaffold an SNS resource so that learners gradually take ownership. For example, Yang, Crook & O'Malley (Reference Yang, Crook and O'Malley2014) examined the implementation of a Ning-based SNS created for seven adolescent learners in an extracurricular Mandarin learning program for virtual teaching, social, and cognitive presence. Over time the teacher gradually relinquished activity direction to the students, who then autonomously assumed teaching presence. Analysis of user behavior and change in content over time showed that the SNS activity exhibited social and cognitive presence as well. The results contrast with Kennedy & Miceli's findings (Reference Kennedy and Miceli2013, §3.1.3), where a learner-directed wiki-enhanced learning resource was not so successful—it may be that without scaffolding, or perhaps for larger numbers of students taking a course as a requirement rather than voluntarily, SNS resources will not be utilized as readily for autonomous learning.

A third approach might be to offer a variety of platforms and allow learners to choose and use them differently. Lamy (Reference Lamy and Meskill2013) examined about 5400 SNS-enhanced posts by L2 Chinese learners who, over a year, participated in two institutional but informal chat rooms (4000 posts) as well as in public and private Facebook groups they created for themselves (1400 posts), posting freely in both languages. She found that most learner posts, in both the institutional and Facebook groups, exhibited re-use and link display, but little openness or network effects—the former meaning sharing resources that are openly re-useable, and the latter meaning enhancing shared resources so that they grow in virality or popularity—all four qualities can lead to affordances. She attributes differences in post type to the differing parameters and constraints of participation in the institutional, informal, and learner-directed groups. It may also have been that particular site features facilitate or inhibit open or network activities, or that even in the informal groups the learners saw the activities as formal in nature. Implications from research on informal uses (§4.1) might be applied, and more long term research on the use of semi-formal SNS groups is warranted—perhaps using ethnography to get at users’ emic perspectives on what they believe they are doing in the different SNSs, and for what purposes they use one kind over another.

4.2.6 Summary and future directions: SNS-enhanced L2 pedagogy

In sum, research has shown that vernacular SNSs can be used for observation of, and participation in, socio-pragmatically genuine discourse practices in the L2 of study. The SNS practice of profile curation can be leveraged as an affordance for L2 identity development, interaction, and language awareness through simulation and role play activities, while network traversing (i.e. sharing and posting) can be leveraged to develop learner community. Learner agency and autonomy can be facilitated through the provision or unofficial support of SNS resources that are supplemental to formal instruction. Findings also suggest, however, that resistance may be encountered depending on task parameters and learner variables—for example if SNS use is required, if activities are imposed top-down, and if learners are required to use their real identities and to friend each other unwillingly.

Future research and instructional development should build on the findings from analysis of naturalistic, informal SNS L2 use and learning (e.g. § 4.1.4), connecting descriptive research with pedagogical practice. As SNS use becomes everyday practice and assumes a central role in our cultural, political, and educational lives, development of social media literacies—by using SNSs critically as windows, gateways, and playgrounds—might be a direct goal of L2 curricula. Through critical examination of how languages and other semiotic means are used in SNSs for presenting identities, building and maintaining relationships, and otherwise creating social meaning, learners might see their own and others’ practices more critically, and participate more reflectively in both L1 and L2. Autonomous use is key to successful lifelong L2 learning—there is no doubt that once L2 learners leave classrooms, there is a good chance they will encounter the L2 in social media and other digitally enhanced contexts as often, if not more often, than in face-to-face situations.

4.3.1 Overall SNECS usability

Research focused on SNECS usability has generally found that while different features appeal to different learners, poor design often distracts from learning potential. For example, Harrison & Thomas (Reference Harrison and Thomas2009) followed six L2 learners who used Livemocha. Using boyd & Ellison's (Reference boyd and Ellison2007) conceptualization of SNS identity as an analytic framework, they argued the site afforded self-presentation, network management, and community participation, but that different site features appealed to some learners but not to others. Clark & Gruba (Reference Clark, Gruba, Steel, Keppell, Gerbic and Housego2010) reported on their own experiences using Livemocha to learn Korean over four weeks, identifying several site affordances that motivated them and aligned with L2 pedagogical principles—the opportunity to communicate with native speakers, to use chat instead of voice or video, to leave comments for others, and to have tasks broken into manageable chunks. However, certain design elements led to frustration, including outdated methods and boring drills, decontextualized content, limited opportunities for interaction, excessive requirements to help others, and no acknowledgement in the site design of the relative difficulty between English and languages like Korean as opposed to Spanish.

The work of Liu and colleagues (Stevenson & Liu Reference Stevenson and Liu2010; Liu et al. Reference Liu, Evans, Horwitz, Lee, Mccrory, Park, Parrish, Lamy and Zourou2013; Liu et al. Reference Liu, Abe, Cao, Liu, Ok, Park, Parrish and Sardegna2015) also employs usability testing along with comparison as methods for evaluating SNECSs and their potential integration into formal instruction, with similar critical findings. Comparing Palabea, Livemocha, and Babbel, Stevenson & Liu (Reference Stevenson and Liu2010) evaluated technical and pedagogical usability by examining how five learners completed exploratory, open-ended, and close-ended tasks in the sites. In brief, users wanted a combination of features no one single SNECS offered—access to native speakers and tutorial CALL resources, an atmosphere of both serious and fun learning, customizability, and intuitive interface design. In a second study, Liu at al. (Reference Liu, Evans, Horwitz, Lee, Mccrory, Park, Parrish, Lamy and Zourou2013) explored the possibility of integrating SNECSs into formal instruction. They had 21 adult ESL students complete a series of focused tasks over six weeks in Livemocha, Busuu, and English Café, and found in general that the first was preferred for speaking and vocabulary, the second for reading and writing, and the third for grammar. There were considerable differences in how comfortable participants were using the sites, which communication modes they preferred, how proficiency level influenced outcomes, how feedback was received, and how design features led to preferential differences because of perceived affordances. A third study (Liu et al. Reference Liu, Abe, Cao, Liu, Ok, Park, Parrish and Sardegna2015) examined four more SNECSs—LingQ, Lang-8, italki, and Polyglotclub—and focused on the design of supplemental learning tasks they designed, including journal keeping, video watching, text reconstruction, and photo sharing tasks. Behavioral data analysis and a user survey of six adult ESL learners revealed that advertisements, complicated and poorly designed interfaces, and inaccurate translations sometimes interfered with the tasks. In short, teachers may not be able to ameliorate the problems of poor SNECS design with well-designed supplemental tasks.

Coming to similar conclusions, Brick and colleagues (Brick et al. Reference Brick2011; Brick et al. Reference Gibson2012; Orsini-Jones, Brick & Pibworth Reference Orsini-Jones, Brick, Pibworth, Zou, Xing, Xiang, Wang and Sun2013) also examined SNECSs as potential formal teaching resources, from the perspectives of both learners and instructors. Brick's analysis (Reference Brick2011) of the Livemocha experiences of seven learners of a variety of languages showed frustration at the quality of peer feedback as well as a perceived tendency towards ‘cyber-flirting’. Users were quite unhappy with the quality of the curricula, which focused on rote memorization and word lists. In a complementary study (Reference Brick2012), Brick analyzed the experiences of 15 instructors and 14 students with Busuu. While most reported their overall experiences were positive, and students were enthusiastic about the site's reward systems, there was some instructor skepticism about inauthenticity of material and cultural representations. Finally, Orsini-Jones, Brick & Pibworth (Reference Orsini-Jones, Brick, Pibworth, Zou, Xing, Xiang, Wang and Sun2013) examined how eight L2 teacher trainees evaluated Livemocha and Busuu with regards to learner agency and found a pattern of motivation, frustration, and demotivation similar to what Clark & Gruba (Reference Clark, Gruba, Steel, Keppell, Gerbic and Housego2010) found. While participants appreciated some site features, they were demotivated by mismatches between materials and designated level and between the potential of the sites to support authentic interaction and the behaviorist, decontextualized nature of the materials themselves. They were also frustrated at the lack of pedagogical expertise by the other users, even if they were recognized by the site to have it.

Most recently, large scale quantitative usage analysis offers statistically significant evidence that lack of real teaching expertise among users is perhaps the biggest issue with SNECSs. Lin, Warschauer & Blake (Reference Lin, Warschauer and Blake2016) surveyed over 4000 users of Livemocha, with additional case study analysis of the production of 20 users, focused on attitudes and site usage. Although survey participants in general felt the site built self-confidence and was motivating, the case studies showed that most users quit after a few months. Analysis of user work showed that while syntactic complexity increased the longer the site was used, accuracy did not. The researchers argue the findings implicate the need for dedicated, trained instructors, without whom learners are not motivated to persist in their efforts and to attend to persistent errors in accuracy.

4.3.2 Focus on specific SNECS features and affordances

A second strand of research on SNECSs focuses on specific site features and their affordances, often taking theoretically rigorous discourse analytic approaches to do so. This approach, to focus on features and dynamics that are universal across SNECSs, is perhaps in reaction to the fact that as commercial products being constantly updated, SNECSs are evolving and sometimes disappearing targets. Most studies, like those employing usability analytic methods, find that site designs leave users frustrated, and that while negative affordances counteract positive learning outcomes, sometimes positive affordances emerge in spite of site design, as users adapt.

For example, Harrison (Reference Harrison, Lamy and Zourou2013) examined the central role that profiles play in impression management and identification of the pedagogical expertise of language partners in SNECSs. Ethnomethodological analysis of the experiences of seven Livemocha users found they were discouraged by the inability to control profile features and privacy settings, which, combined with the inability to discern authenticity in others’ profiles, led to mistrust and uncertainty about the potential affordances of peer interaction and tutoring. In a similar focused study, Zourou & Loiseau (Reference Zourou, Loiseau, Lamy and Zourou2013) conducted a quantitative analysis of the Livemocha's culture section's threads over two years, showing a boom at launch of number of threads (about 1300 per day), a gradual decline of threads over time, and a low rate of persistence—only 10% remained active after a month. Closer analysis found that most threads were photos, comments were very sparse, language used was simple, and responses on the most popular threads were not interactive. The researchers conclude that the mechanism for sorting threads favored new thread authoring rather than response or reuse and that the functional complexity of network traversing—i.e. finding and interacting with one's friend's culture threads—inhibited the potential of the culture section to afford social interaction and learning. Most recently, Zourou, Potalia & Zourou (Reference Zourou, Potolia, Zourou, Lewis, Rivens Mompean and Cappellini2017) explored SNECS affordances for the development of autonomy in a study of over 1500 French and English learners using Busuu. Comparison of the users’ solo engagement with tutorial CALL exercises to social activity with peer correction activities showed a majority preference for the latter, which the authors argue represents a social autonomy stance. Whether or not that activity leads to long term learning gains, however, is unknown—although if Lin et al.’s (Reference Lin, Warschauer and Blake2016, §4.3.1) findings hold, it probably does not.

Gonzales (Reference Gonzales and Meskill2013) used conversation analysis to examine how an L2 Spanish learner negotiated rapport and solidarity in Livemocha chat with native-speaking partners in spite of, not thanks to, site affordances, through the use of humor, small talk, textualized paralanguage, and shared cultural reference. Even though the site provided chat users with translators, extra emoticons and L2 characters, and suggested topics, the learner instead autonomously leveraged his CMC discourse competence, thereby developing sociopragmatic competence. In a similar study using discourse analysis, Vandergriff (Reference Vandergriff, Dixon and Thomas2015) examined 120 discussion board transcripts in the non-commercial, non-profit SNECS Deutsch für Dich (DfD) for evidence of the expression and exercise of learner agency, including formulating goals, engaging in strategic learning behaviors like planning and self-regulating, choosing which language to use, and engaging in humor. She found considerable evidence of the expression of agency, sometimes, like Gonzales (Reference Gonzales and Meskill2013) in spite of site design, but mostly in concord with the pedagogical objectives built into the design. It is notable that DfD is developed by the venerable non-profit Goethe Institute, which hires professional tutors and trained instructors to operate and design the site.

4.3.3 Summary and future directions: SNECSs

In sum, most studies that have examined SNECSs in terms of overall usability by learners and instructors have found that poor site design not grounded in best L2TL practices or accepted SLA theory ultimately leads to frustration and abandonment. Studies of particular site features and affordances emergent from use have also shown less than encouraging results, although recent studies (e.g. Vandergriff Reference Vandergriff, Dixon and Thomas2015; Zourou et al. Reference Zourou, Potolia, Zourou, Lewis, Rivens Mompean and Cappellini2017), aligning with trends in other social media research (cf. §2.1.3, §4.1.2, §4.2.5), show some affordances for the development of learner autonomy.

While the evidence against Livemocha might lead one to conclude its demise was perhaps for the better, other SNECSs continue to develop, and one hopes their developers might learn from the successes of the non-commercial DfDs among them. Unfortunately, as long as there is a steady stream of new customers who find the status, ranking, and gamified rewards systems motivating, but who are unaware that poorly designed, decontextualized memorization drills may not lead to long term learning gains, commercial SNECS developers have no reason to pay attention to L2TL research. To force their and their customers’ attention, more long term studies like Lin et al.’s (Reference Lin, Warschauer and Blake2016) that target and isolate different features, mechanics, and dynamics are necessary. The question of whether sustained, long term use of commercial SNECSs can lead to sustained linguistic development, even in the absence of teaching expertise, can be answered with studies that compare SNECSs with other SNECSs as well as with non-SNECS conditions. In particular, qualitative studies of successful users and formal programs that integrate SNECSs are needed that offer insights into the ecologies of SNECSs use.

Increasing research focus on informal, extramural learning practices may also contribute to research on SNECSs and social media in L2TL more broadly (e.g. Sockett Reference Sockett2014; Sylven & Sundqvist Reference Sylven and Sundqvist2016). While educational social media use is only one kind of practice among a variety of informal, self-directed L2 learning practices, SNECSs are designed for intentional L2 learning, unlike digital games or other vernacular media. Chik & Ho (Reference Chik and Ho2017) found that beginning L2 learners gravitated towards structured self-directed L2 learning resources rather than opportunities for interaction with other speakers, and that time limitations pushed would-be learners to ready-made rather than do-it-yourself resources. While some SNECSs become more comprehensive as others become more targeted and customizable, their place in an L2 learner's increasingly multifarious repertoire of extramural resources is a topic for continued scholarly attention.