Social Media in Second and Foreign Language Teaching and Learning: Blogs, Wikis, and Social Networking
Social Media in Second and Foreign Language Teaching and Learning: Blogs, Wikis, and Social Networking
This review surveys and synthesizes the findings of 87 focal pieces, published primarily
between 2009 and mid-2018, on the formal and informal use of social media—blogs,
wikis, and social networking—for second and foreign language teaching and learning
(L2TL), including studies on the use of educational sites like Livemocha and Busuu and
vernacular sites like Facebook and Twitter. The article frames the review in the
development of social media and the history of social computer-assisted language
learning research. Synthesis identifies common findings, including that social media can
afford the development of intercultural, sociopragmatic, and audience awareness,
language learner and user identities, and particular literacies. Presentation of the focal
pieces and common findings is intertwined with discussion of problematic issues, and
each section concludes with a summary and implications for future research and practice.
1. Introduction
The Internet has greatly impacted language teaching over the past 20 years, as evidenced
by a substantial and growing body of research in a variety of fields, including language
pedagogy and assessment, second language acquisition, discourse analysis, literacy
studies, computer-mediated communication, and sociolinguistics. The field of computer-
assisted language learning (CALL) in particular has concerned itself with the potential of
the Internet for language learning, starting with the publication of several seminal articles
in the mid 1990s that touted the benefits of computer-mediated classroom discussion (e.g.
Chun 1994; Kern 1995) and intercultural exchange (e.g. Warschauer 1996). While CALL
today spans a range of issues as diverse as online learning, computer adaptive testing,
digital gaming, and corpus linguistics (see recent handbooks edited by Farr & Murray
2016; and Chapelle & Sauro 2017), there has been increasing interest in the social aspect
of Internet use since SLA has made the ‘social turn’ (Firth & Wagner 1997; Block 2003)
and ‘Web 2.0’ technologies, along with expanded broadband and mobile accessibility,
started supporting new ways of social interaction (Lankshear & Knobel 2006;
Warschauer & Grimes 2007). Beginning in the mid 2000s, technologies based on Web
2.0 platforms became popularly known as ‘social media’, including blogs, wikis, social
networking and a variety of related platforms, services, and media technologies.
To conduct the review I followed Norris & Ortega’s (2006) procedures for
conducting a research synthesis: 1) identify studies and selection criteria; 2) focus on
variables and data and draw one’s own conclusions, rather than researcher conclusions;
3) examine parameters and categories that cut across studies; and 4) establish super-
ordinate heuristics to synthesize conclusions. To identify studies I searched for associated
keywords (e.g. social media, Web 2.0, blogs, wikis, social networking, Facebook,
Twitter, etc.) in the 2009 to mid-2018 issues of the most well-known journals in CALL:
Language Learning and Technology, CALICO Journal, ReCALL, Computer Assisted
Language Learning, and System. In addition, I searched the aforementioned edited
volumes and special journal issues for key studies prior to 2009. Finally, I checked
bibliographies of the most recent articles and chapters for pieces that may have appeared
in other journals or volumes. I read 160 pieces to identify research strands into which
multiple studies could be categorized over several of the target years. I then chose
empirical studies that I judged to be representative of the identified strands as focal
pieces, as well as a few pedagogical or ‘idea’ pieces that I felt useful to frame a particular
research strand, totaling 87 pieces. While some of the pieces reviewed under these
criteria have low participant numbers or are qualitative in nature, and therefore are not
necessarily replicable or scientifically generalizable, they are still important to consider
because they are empirically grounded and situated in ecologically valid contexts of
implementation. From a phenomenological perspective, consideration of such studies en
masse can perhaps better inform pedagogical theorizing and reflective practice because
their implementational contexts are more relatable than those of large experimental
studies.
The review is divided into three sections: blogs (§2), wikis (§3), and social
networking sites and services (SNSs) (§4), which is subdivided into three sections:
informal SNS-enhanced L2 use and learning (§4.1), SNS-enhanced L2TL pedagogy
(§4.2), and L2TL with SNECSs (social network enhanced commercial CALL sites and
services) (§4.3). There was not enough published empirical research on L2TL with other
social media to identify research strands and trends and thus warrant inclusion here.
Blogs, wikis, and SNSs are the most researched because they have been around the
longest and are most popular, a fact that in no way discounts the necessity for research
Since long before inventing the Internet, humans have developed technologies, from
writing to calling cards to the telegraph with which they engage in social networking,
information sharing, and joint creative activity (Standage 2013). In the Internet’s early
days, Usenet and bulletin board systems supported asynchronous discussion and resource
sharing, and graphical webpage software and hosts afforded the creation and hyperlinking
of personal homepages. Early Web-based services like Compuserve and America On Line
offered users a single interface for accessing news, sharing information, and interacting
socially with other network users. In the mid to late 1990s, authors of online personal
journals began hyperlinking their blogs (from ‘weblogs’) with others in what became
known as the ‘blogosphere’. Wikis (from the Hawaiian wiki-wiki for ‘quick’) first
appeared at about the same time as quick means for programmers to share and organize
resources collaboratively. In the mid 2000s, Friendster, MySpace, LinkedIn, and
Facebook launched social networking sites that allowed individuals to create profiles, to
connect to profiles of other users they knew or wanted to know in real life, and to share
content with them. Web services that focused on media creation, curation, and sharing,
like Flickr and Youtube, gradually integrated social networking elements and by the
2010s became understood as ‘social media’.
Some estimates are that as of 2018, nearly a third of our planet’s population, or
2.6 billion people, make some use of social media (Statista 2018). This massive increase
in usage is partially due to the ‘Web 2.0 turn’ in the mid 2000s along with the continued
development of affordable, portable, and accessible digital information and
communication technology. Coined as a retronym, the term Web 2.0 refers to the second
generation of World Wide Web technologies that allowed more user participation and
content sharing than was possible with older Web 1.0 technologies, leading to the
profusion of social media. What we recognize as social media today has roots in the
technologies of computer-mediated communication (CMC), personal homepage
authoring, and cyber-community participation, but it was the capacity of new Web 2.0
technologies starting in the early 2000s to separate content from form that allowed for
users to produce content and articulate networks, giving rise to blogs, wikis, and social
networking sites.
As long as meaningful social interaction has been understood as necessary for language
learning, language educators have sought to integrate a means for communicative
practice into teaching. Some of the first Internet applications for synchronous (chat) and
asynchronous (email and BBS) communicative exchange were identified as having
educational potential, and teachers began implementing them, and researchers
investigating them, as soon as it was logistically feasible. Early empirical research (e.g.
Beauvois 1992; Chun 1994; Kern 1995; see Ortega 1997 for a summary) found evidence
that the various features of CMC led to increased output production, access to a wider
range of discourses, equalized participation, and increased fluency—albeit sometimes at
the expense of accuracy. These potential affordances are still associated with use of social
media in L2TL.
In their research designs, researchers of the social Internet found cognitive and
psycholinguistic frameworks insufficient to account for some of the outcomes of
technology-mediated discussions and telecollaborations (see Reinhardt 2012). Inspired by
the ‘social turn’ in SLA (Firth & Wagner 1997; Block 2003), many turned to socially-
informed theoretical frameworks—for example, sociocultural theory, situated practice,
language socialization, or socio-cognitive frameworks that view language development as
an essentially social phenomenon. In the introductory chapter to their seminal 2000
edited volume Network-based Language Teaching, Kern & Warschauer explained that:
In other words, well before Web 2.0 and popular conceptualizations of social media, L2
educators and researchers found in socially-informed frameworks theoretical grounding
with which to leverage the socio-interactional affordances of Internet-mediated
communication—email, discussion boards, and chat—for the purpose of language
teaching. The potentials for language use and learning, or ‘affordances’ (Gibson 1979;
van Lier 2004), of these tools allowed individual expression through homepage authoring
and participation in ‘cyber-communities’ not bound by physical and temporal immediacy.
2. Blogs
A blog, short for ‘weblog’, is a journal-like website comprised of dated posts, presented
in reverse chronological order, often with threaded comments under each post. Blogs
offer ‘a mixture in unique proportions of links, commentary, and personal thoughts and
essays’ (Blood 2000) and are thus argued to be ideal spaces for the development and
expression of expertise. Blogs are linked to other blogs by their authors, thus creating
networked communities of writers and readers, known collectively as the ‘blogosphere’.
Several genres have emerged over time (Herring et al. 2005, in Warschauer & Grimes
2007), including personal blogs, filter blogs, and knowledge blogs—roughly equivalent
to online diaries, news journals, and topical guides, respectively. While genres differ
according to purpose, audience, and author role, blogging foregrounds the act of
individual, nonymous authorship (as opposed to anonymous—see Zhao, Grasmuck, &
Martin 2008), although some genres, especially filter blogs, allow for multiple authors.
Blogs were arguably the first social media because they were designed to support
interactive readership and multimedia embedding from their inception in the late 1990s.
Once Web 2.0 allowed individuals with no knowledge of coding to create blogs, the
practice boomed—WordPress, only one of many blog hosts, published over 18 million
blogs in 2014 alone (Crum 2015). More recently, however, as interactive personal Web-
based writing has become integrated into social networking and other social media like
Twitter (which is sometimes called ‘microblogging’—see §4), traditional, single-authored
blogging may be waning in general popularity and becoming a niche practice.
Still, as has been noted by many L2 educators and researchers, blogging may
continue to hold potential for language learning because of its unique affordances for
writing as social and reflective practice. Early pieces were often speculative and
exploratory, describing potentials not yet empirically examined. Godwin-Jones (2003),
for example, discussed the Web 2.0 origins of blogs and the potential for blogs to serve as
personal journals or portfolios, because they seemed to support reflective learning and the
development of a sense of ownership or authority. Campbell (2003) offered a framework
of blog types for educational application: tutor blogs as instructor-maintained collections
of class references and resources, learner blogs as online journals or portfolios for
individual students, and class blogs as spaces for collaboration and interaction (cf. Alm
2009, §2.1.3).
Since the mid 2000s, research on blog-enhanced L2TL has followed several intertwined
strands: 1) blogs as media for culture learning and intercultural exchange, 2) blogs as
spaces for literacy and identity development, the purported affordances of blogs for
developing 3) learner autonomy and 4) audience awareness, and 5) the importance of task
design and learner variables, a thread that emerges from and through other strands.
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 (2008) 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
2009, §2.1.3) and issues of task design (cf. Lin, Groom, & Lin 2013; Chih 2015, §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 (2008) 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 (2009) 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 (2012)
combined some of Ducate & Lomicka’s (2008) and Elola & Oskoz’ (2008) task designs,
While methodologically rigorous, the facts that Bloch (2007) and Gebhard et al.
(2011) presented single case studies and that Sun & Chang’s (2012) and Arshavskaya’s
(2017) 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 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.
Another strand of research on blog-enhanced L2TL argues that blog-based tasks should
support and develop learner autonomy. For example, Alm (2009) designed blog-
enhanced German instruction with the goal of enhancing learner autonomy. Her design,
informed by self-determination theory (Deci & Ryan 2007, in Alm 2009: 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 (2009) 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 et al. (2013) 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 8 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.
However, like the aforementioned nuance that Lin et al.’s (2013) work added to
understandings about learner autonomy, more recent research also implicates task design
in purported affordances for audience awareness. To illustrate, Chen (2015) 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 2007). 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 2015,
§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 (2009, §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.
Chen, Shih, & Liu (2015) focused directly on blog-enhanced L2 learning task
design, which they argue may be responsible for some negative findings, like those of
Krause (2005, in Chen et al.: 286), 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.
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 of which the learner has had some choice. Later research has found
that tasks which force particular topics, require form-focused peer review, overly restrict
or broaden audience, or necessitate closed outcomes 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.
Warschauer & Grimes (2007) argue that unlike blogs, which highlight self-
presentation, voice, and individual authorship, wikis seem to promote collaborative,
distributed authorship. To generalize, social networking sites (SNSs) and wikis afford
focus on sharing and networking, while blogs promote, but do not necessarily require,
dialogic interaction and collaboration. Blogs and wikis afford resource production and
idea generation, while SNSs allow for focus as much on sharing across networks as on
original production. Because of their designs, blogs and SNSs seem to highlight
individual authorship, while wikis de-emphasize the identities of individual contributors.
Similar to wikis, collaborative documents like Google Docs have emerged as an object of
CALL research recently (e.g. Kessler, Bikowski, & Boggs 2012; Bikowski & Vithanage
2016), but they are not included in this review because unlike wikis, they are not
normally used for public documents or resources and so do not function as social media.
As with blogs, a main social learning affordance argued by wiki-enhanced instruction
advocates is increased audience awareness, since learners recognize their work is publicly
accessible.
When Web 2.0 technologies made wikis a practically feasible reality in the mid
2000s, new perspectives and speculations on their potential for collaborative writing
began to emerge in the literature (Godwin-Jones 2003), based on the state of L2 writing
research at the time. It focused on the learning potential of computer-mediated peer
review (e.g. Ho & Savignon 2007) using commercially available tools available to a
group of users on a closed, local network—the Daedalus Integrated Writing
Environment, for example, has been available since the early 1990s. Research on
classroom-based collaborative writing was focused on the social potential of writing in
Since 2008, research on wiki-enhanced L2TL has continued and expanded research
strands begun in CMC and collaborative writing circles by focusing on wiki affordances.
Although most studies cover several areas, they can be categorized according to their
focus on purported affordances for 1) meaning over form, 2) collective activity,
collaboration, and cooperation, and serving as 3) virtual learning environments. As with
blogs, wikis are also claimed to support development of 4) awareness of genuine and
imagined audiences, although again 5) task design parameters and learner variables play
key roles.
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 (2009) 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.
Bradley et al.(2010) note critically that ‘the potential of the wiki concept as a
writing tool is frequently assumed’ (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 (2012) 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.
Because wikis are easily customizable, they can be used as virtual learning environments
that provide supplemental resources and scaffolding on how and why to use wikis for L2
learning. For example, deHaan et al. (2012) 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 2012, §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 (2013) sought to use a wiki for their
university level beginning Italian learners at 3 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 (2013) 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
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 (2008),
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; 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. 2012, §3.1.3). For example, Pellet (2012) 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 2009, §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 also account also for the successes of
King’s project (2015), 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
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 (2010) 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 2009,
§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 (2013) 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 show findings similar to Arnold et al. (2012) and Li & Zhu
(2017, §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 (2014) 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 2010, §3.1.1). Moreover, they found half of the students didn’t
enjoy the tasks—perhaps like Chen et al.’s (2015, §2.1.4) blogging students they would
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
should be anticipated. Wikis may serve as virtual learning environments, 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.
Social networking sites (SNSs) can be defined as social media that foreground personal
profile curation, network traversing, and network articulation (boyd & Ellison 2007;
Ellison & boyd 2013)—in contrast to a primary emphasis on content creation, which
typifies blogs, wikis, and other sorts of social media. However, as the technologies of
social networking, the Web, broadband, and mobile and smartphones continue to evolve,
functional distributions among different platforms and tools have blurred. While SNSs
are typified by user status updates presented in reverse chronological order, which
function somewhat like short blog posts with threaded discussions, they integrate
functions like photo hosting, multimedia sharing, chatting or messaging, and a variety of
As SNS use has become global and mainstream, academia has taken interest.
Education scholars (e.g. Selwyn 2008) have argued that SNSs can facilitate the
development of collaborative and participatory learning communities, as well as
opportunities for informal and unstructured learning. In a review of over 400 studies on
Facebook, Wilson et al. (2012) identified five major areas of research focus: descriptive
analysis of users, motivations for using Facebook, identity presentation, the role of
Facebook in social interactions, and privacy and information disclosure. In line with
education, communications, and social science research, L2TL researchers and
practitioners have explored informal L2 use and learning using vernacular SNSs like
Facebook (§4.1), formal pedagogical applications of vernacular SNSs like Facebook and
Twitter (§4.2), and the informal but intentional use of educational SNSs or SNECSs
(social network enhanced commercial CALL sites and services) like Livemocha and
Busuu (§4.3).
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 (2006) 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 2008), and Welsh users in the UK (e.g., Cunliffe, Morris, & Prys
2013; Jones 2015). 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 (2010) 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 (2011) 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).
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 (2011) analyzed the online SNS, blog, and
email-mediated writings of 12 Australian learners of Japanese and their contacts over
four 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
Other studies have employ mixed and quantitative techniques as well. For
example, Chen (2013; see also Chen 2012) 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 (2013) 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 2013). Also employing mixed methods, Solmaz
(2015) used an innovative visualized social network analysis technique to present
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.
Complementing research on informal L2 use and learning with vernacular SNSs, and
sometimes incorporating its implications, a second broad research area on SNS reports on
formal, classroom-based pedagogical applications of vernacular SNSs like Facebook and
Twitter, as well as on autonomous learning through extracurricular use of SNSs
associated with formal interventions. Research strands in the area focus on 1) the
motivational benefits of SNSs, and their affordances for 2) observation of and
participation in genuine L2 socio-pragmatic usage, 3) situated and simulated practice, 4)
literacies development, and 5) autonomous learning.
Early work on vernacular SNS use included idea pieces focused on its motivational
potential. For example, McCarty (2009) 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 2009)—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 (2010) found that 4 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 4 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 2011).
One of the first to propose potential formal applications of SNSs, McBride (2009)
describes them as spaces for learner self-authorship, interaction, and sociopragmatic
development. Echoing Prensky’s (2001) 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.
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 (2009) 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 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 (2011) 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
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
2011; cf. also §3.1.3). These might be integrated with ‘SNS literacies’, which Solmaz
(2015) 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 (2013)
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 (2013) 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 2009; §3.1.1)
Because SNSs are associated with informal, everyday uses, formal SNS activities
may not always be well-received, as Reinhardt & Zander (2011) 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 (2014) 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 (2014) 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.
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.
A third area of research on SNSs for L2 learning is on the self-directed use of SNECSs,
sometimes called social networking sites for language learning (SNSLL). Zourou (2012)
offers a taxonomy for three different types of these ‘Web 2.0 language learning
communities’, comprised of ‘structured language learning communities’, centered on
tutorial CALL lessons, ‘marketplaces’, focused on tutor or peer pairing and hosting, and
‘language exchange sites’, focused on general interaction and socialization. As of this
writing, however, these sites have continued to evolve into online amalgams of all three
types enhanced by a variety of social networking and intelligent CALL features. In
March 2016 the most widely researched SNECS, Livemocha, was acquired and shut
down by Rosetta Stone, which may or may not revive it in some form. Busuu, Babbel,
and Palabea (also now defunct) began as primarily online tutorial CALL and have
incorporated various social network-enhanced market and exchange features from their
beginnings. DuoLingo started as a peer-tutoring translation site, and has increasingly
incorporated tutorial CALL elements and social networking features.
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 (2009) followed six L2 learners who used Livemocha.
Using boyd & Ellison’s (2007) 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 (2010) 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 (2010; 2013; 2015) 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 and Liu (2010) 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. (2013) 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
Coming to similar conclusions, Brick and colleagues (2011; 2012; 2013) also
examined SNECSs as potential formal teaching resources, from the perspectives of both
learners and instructors. Brick’s analysis (2011) of the Livemocha experiences of 7
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 (2012), 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 (2013) examined how 8 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 (2010) 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.
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.
For example, Harrison (2013) 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 7 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 an similar focused study, Zourou & Loiseau (2013) 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 re-use 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, and
Zourou (2017) 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 (2016, §4.3.1) findings hold, it probably does not.
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 2015; Zourou et al. 2017),
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 (2016)
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.
5. Conclusion
Research on social media in L2TL to date has utilized socially informed L2TL
theories, especially social-constructivist, multiliteracies, and L2 identity development
theories, as well as a variety of research methodologies and techniques, including case
studies, instructional and action research, discourse analysis, ethnography, and
quantitative data analysis. As practices mature, they have the potential to inform theory
building and methodological innovation in the fields from which they have borrowed,
including computer-mediated communication, digital humanities, new media studies, the
learning sciences, and educational technology. Challenges include the constraints of
technology, the limits of privacy, and reconciling the formal demands of research with
the informal, anywhere, any time qualities of social media use.
It should not be overlooked in either research or practice that all vernacular social
media, and most all educational social media, especially SNECSs, are usually
commercial enterprises. As such, they should be subjected to the same critical scrutiny as
any media used for L2TL as a means or object of learning. Advanced learners especially
can benefit from explicit media literacy instruction (e.g. Buckingham 2003) applied to
social media, with discussion of different media types and sources in the L2, perhaps in
comparison to the home culture/language (Reinhardt & Thorne 2011). The concepts of
user agency, control, and choice—especially in situations where the user is involved
through Web 2.0 mechanisms as a producer of content—should be critically situated and
understood in terms of how language is used in and by social media agents, for what
References