Language Education in Emergencies: A Systematic Review

In response to the Covid-19 pandemic, education systems around the globe suspended onsite classes and transitioned instruction to various remote environments, creating a distinctive context for teaching and learning. This systematic review brings together the available research in language education within the current emergency setting to examine the state of affairs, as well as the situation’s inherent challenges and opportunities for language learners and teachers. A total of 38 studies were collected to reflect the current trend, with 16 of these indepth reviewed. Research focusing on learners was dominant. Most research was conducted at the tertiary level. The studies highlighted digital tools that are capable of engaging language learners in an interactive learning experience, though they are unable to replace face-to-face instruction. Outside-of-class support such as extra channels of communication, self-access language learning (SALL) materials, and advising in language learning (ALL) were all found to complement remote learning. It is recommended that teachers try to retain their teaching principles and put them into practice regardless of the abrupt transition. Teachers’ wellbeing can be promoted when teachers accept the changes and see them as opportunities.


Introduction
Late 2019 saw the greatest challenge humankind has faced in its history: the outbreak of a novel coronavirus , which was shortly later characterised as a pandemic. At the time of writing, there have been 111 million confirmed cases, with the loss of 2.46 million lives. As well as health and economic crises, the pandemic has also caused the largest disruption to education systems in the world's history, affecting 1.6 billion learners in more than 190 countries 1 . Educational institutions were temporarily closed in an attempt to minimise the adverse impact of the pandemic, bringing about a transition in the mode of teaching delivery from on-site to remote instruction that, in most cases, relied heavily on digital resources. Various efforts and arrangements were made to support the continuity of education. Classroom teachers all over the world, despite the very limited experience with remote education, have been dedicated to maintaining access to learning for all students.
In language education, engaging learners in the sudden transition to remote teaching has been a challenging task (Gao & Zhang, 2020;MacIntyre et al., 2020;Moser et al., 2021;Thumvichit, 2021). The pandemic has created a distinctive context for the remote teaching of language (Moser et al., 2021). Though distance education has existed for centuries, implementing it without adequate time for preparation, warning, resources, and the implementation has been far from simple. The intersection between the Covid-19 crisis, the abrupt shift to remote education, and language education in particular can shed light on future research directions and practice in difficult times. This review is designed to create a reference point, generating fresh insights into language education in emergencies for educational leaders, teachers, students, and researchers. It is important to note that conducting a review of research in a developing disciplinary area is more complicated than conducting one in a "mature" disciplinary area (Li et al., 2020). Language education in emergencies is not yet well defined, and therefore conducting a review of this area requires careful thought and a clear scope to handle the complexity. This review covers a range of both L1 and L2 educational contexts.

The Abrupt Transition to Remote Teaching
Digital tools for teaching and learning are not new, but using them in restricted conditions (e.g., with resource constraints, time limitations, information asymmetry) poses unprecedented challenges. Teachers have struggled to find methods that can help them sustain education and keep their students engaged during the suspension of face-to-face classes. The difference between the types of distance and online education that we have been familiar with for a long time and the abrupt transition to remote teaching is that the former usually includes a sufficient preparation period. In contrast, the latter is a short-term and sudden solution to education during calamities such as natural disasters, wars, epidemics, and so on (Moser et al., 2021). The latter type of shift occurs 'when teachers and students are not accustomed to using online platforms and technology' (Russell, 2020, p. 339). Emergency remote teaching can take on various forms, including radio or television broadcasts, online platforms, and mailed learning materials. Emergency remote teaching will never share characteristics with planned distance education (Moser et al., 2021). While the function of distance education is to provide access to learning when learners and information sources are separated by time and/or distance (Nwezeh, 2011), emergency remote teaching is about creating short-term access to learning while constrained by a crisis, rather than (re)creating an instructional ecosystem 2 . Distance education is characterised by the flexibility it offers to students, including the choice of modality (e.g., fully online, 50% online, 25% online), place (on campus, at home), and pace (self-paced, class-paced). Emergency remote teaching, on the other hand, does not provide as much freedom as distance education does because it is constrained by national and institutional policies. For example, some teachers are asked by their institutions to record lessons and post them on the institutional platform so that students without access to the internet can study when the opportunity arises.
Forced remote teaching has considerably affected language teachers' ideas about education and language pedagogy (Gao & Zhang, 2020). Since language learning is by nature a discipline in which students are expected to interact with their teachers and peers, engaging students at a distance is inevitably challenging. Teachers' experiences may vary according to context. Resource-rich contexts allow a pleasant learning curve for obtaining technology skills and adjusting teaching approaches 3 . Conversely, in limited-resource contexts, teachers may find themselves struggling for alternative modes of communication and instruction. Although the interplay between the pandemic and language education has received a large amount of research attention, the research community still lacks sufficient knowledge to build an understanding of the overall picture of this event. We posit that emergency remote teaching deserves investigations in its own right, and that it should not be stereotyped as ordinary distance learning.

The Closures of Educational Institutions
Educational institution closures are a common response to crises. Previous discussions on the closures resulting from severe weather conditions or natural disasters provide some useful information relevant to the impact of the school closures caused by Covid-19. Like severe weather conditions and natural disasters, the pandemic took everyone by surprise, and hence disrupted scheduled classes. Closures due to natural factors are analogous to those due to the pandemic (Kuhfeld et al., 2020). Many US studies reported that sudden closures of educational institutions have negatively affected student achievement (e.g., Goodman, 2014;Hansen, 2011;Sacerdote, 2012). Learning loss caused by Covid-19 may be predictable. In a similar previous case of the Ebola crisis, the estimated number of lost learning hours per child was enormous, especially in the epicenter countries. The follow-up impact also cannot be ignored. For example, there was a significant reduction in attendance after schools reopened, reaching levels as high as 25% in Liberia (UNDG, 2015).
However, in the context of Covid-19, education systems put remote education plans into action instantly rather than waiting to resume on-site instruction. Responding to a Gallup survey, over 80% of parents in the US said that their children had access to learning during the pandemic 4 . The remote learning delivered during the Covid-19 lockdowns has helped to offset learning loss. This, however, does not imply that at-home learning can replace on-site learning at educational institutions. One reason is that delivering profound remote learning remains a daunting task. Promisingly, Education Next 5 reported on the results of a survey conducted by the American Enterprise Institute showing that only 20% of US district schools underperformed in relation to their benchmarks after switching to remote instruction. At the same time, students tend to spend much less time on their studies than they did before the lockdown. Some research efforts have addressed factors potentially related to such issues, for example student socioeconomic status 6 and lack of supporting plans (Lake & Dusseault, 2020). Based on data from the Multiple Indicator Cluster Surveys 6 (MICS6), Conto et al. (2020) concluded that missing school during the pandemic damaged children's learning of fundamental skills. Kuhfeld et al. (2020) conducted a prospective analysis based on data from 3 rd -8 th graders, projecting that students would return to schools with around 30% less progress in reading compared with that achieved in a normal school year.
Our primary aim is to document, analyze, and synthesise empirical literature on language education within the context of Covid-19. We intend to address the following review questions: 1. What are the trends in language education research within the Covid-19 context (which areas and educational contexts have been researched)? 2. What are the challenges and opportunities in emergency language learning? 3. What are the challenges and opportunities in emergency language teaching?

Methodology Design
In this review, we adopted a systematic methodology to examine a corpus of empirical research on language education conducted in the context of Covid-19. Although various terms have been used to refer to scholarly reviews: "review", "narrative review", "meta-analysis", "systematic review", and so on, they are to some extent different from one another (Rose et al., 2018). "A confusion of indistinct and misapplied terms" may occur when a review does not fulfil the strict criteria of the selected methodology (Grant & Booth, 2009, p. 91). A systematic review is defined as a review adhering to "a set of scientific methods that explicitly aim to limit systematic error (bias), mainly by attempting to identify, appraise and synthesise all relevant studies" (Petticrew & Roberts, 2006, p. 9), and is characterised by a methodical and replicable process, often involving an exhaustive literature search, the integration of search results, and the synthesis of evidence to address a question (Siddway et al., 2019). Macaro et al. (2017) proposed five features that can be used as indicators of a systematic review. Those features are as follows:

•
Being conducted by more than one reviewer • Using transparent procedures • Including studies through an exhaustive and reliable search process • Minimizing bias to the greatest possible degree • Producing syntheses that contain vivid discussions about the reliability of the evidence We adopted these features to ensure that the current review satisfied all of the criteria of a systematic review.
As far as individual bias is concerned, the authors of this review are a team of three scholars, all from the field of language education with different scholarly interests. The review process consisted of four stages: searching the literature, developing and applying criteria for inclusion and exclusion, grouping qualifying work into research areas and contexts, and conducting an in-depth review.

Literature Search
The literature search was conducted on February 6 th -8 th , 2021, following approval from the Institutional Review Board on January 28 th , 2021. Therefore, articles published during that short time window may not be included. The search was conducted via selected electronic databases: Academic Search Ultimate, Education Source, ERIC, SCOPUS, and Web of Science (Core Collection). These databases were chosen because of their accessibility and worldwide recognition. We also searched mainstream publishers' websites (e.g., De Gruyter, Sage, ScienceDirect, Taylor & Francis Online), on which articles commonly appear before being listed on abstract and citation databases. Different search strategies were applied to ensure that results were as conclusive as possible. In the initial stages, we followed Macaro et al. (2017) evaluating the retrieved abstracts and trying out various search terms. We decided to use broad terms rather than more specific terms, even though it required more screening work. The terms were "language", "additional language", "foreign language", "second language", "L1", "L2", "language teaching", and "language learning", in combination with "Covid-19", "pandemic", "crisis", "closure", "lockdown", and "emergency". Possible synonyms and other parts of speech were applied in addition to the main terms. The search period was limited to articles published between 2019 and 2021. Duplicates were removed by using EndNote.

Inclusion and Exclusion Criteria
The corpus then underwent a screening process in a straightforward manner, focusing only on actual research, not re-hashing calls for change (Rose et al., 2020). To be included in this review, the article needed to: • Report empirical research • Contribute to the effort to address the disruption in language learning brought about by the Covid-19 pandemic • Be published in an academic journal • Be written in English Adapted from Svensson et al.'s (2008) categorisation of publications, the term "empirical research" in the current review refers to studies that include original data. These studies are often further categorised into qualitative, quantitative, and triangular based on their research methodologies. Empirical papers contain methodology sections describing how the data were collected. Conversely, non-empirical papers report on literature reviews, commentaries, researchers' own perspectives, and so on.
All of the criteria were accepted and understood by all authors. Throughout this process, Rayyon (Ouzzani et al., 2016)-a web application for systematic reviews-was used to enable collaboration among authors (see Figure  1). Two rounds of filtering were conducted. First, we worked together on titles and abstracts. Second, each of us worked independently, deciding whether to include or exclude each article. In the case of uncertainty, the article was saved for discussion. Any disagreements were resolved by discussion to the point of consensus. After this process, a total of 38 articles were compiled that met all of the criteria. It is worth mentioning that we identified many more relevant articles but many of them were excluded because they did not report empirical research.

Data Analysis
To address the first review question, we categorised the articles (n = 38) into different areas of focus using open coding (Strauss & Corbin, 1998), as this method allows initial concepts to emerge from data. First, we established research focuses for each article based on its central purpose. This often required us to go through the whole document, rather than merely examining titles and abstracts. After considerable discussion, we agreed on three categories: Learner, Teacher, and Impact of instruction. An experienced scholar in language education who did not feature in the current review was invited to participate in an inter-rater reliability analysis. A short meeting with the external coder was held to clarify the boundaries of each category and explain the procedure. Following the coding, reliability among coders was examined using two standard indices of exact agreement: Cohen's Kappa coefficient (κ) (Cohen, 1960, as cited in McHugh, 2012 and percentage of agreement (PA). The calculation yielded a κ value of .83, which is considered "almost perfect" (Altman, 1999;Landis & Koch, 1977), and a PA of 89.5%. As the first review question also concerns research contexts, we recategorised the articles by context, meaning the educational setting in which the study was situated or the participants' educational stages, ranging from early childhood to tertiary education.
To address the review question 2 and 3, we performed an in-depth review of 16 selected articles (see Appendix). These articles were selected on the basis of relevance to the review questions and readability. A data extraction grid was used, as this allows closer examination of research contribution (Rose et al., 2018). We must note that the current review differs from many systematic reviews in that it does not pay close attention to a single subfield; instead, it reports on what research in emergent subfields has to offer to language education during Covid-19. In this sense, the relevant information synthesised here can be drawn upon in each research area for a deeper insight into the state of affairs in emergency language education.

Risk of Bias
Although a well-conducted systematic review is known for its minimisation of data search and selection bias, our systematic review was somewhat at such risk. First, our literature search was limited to databases that presented only published articles. We acknowledge that there were unpublished works available in other sources, such as conferences, research agencies, and websites. Missing these works may result in the "file drawer problem" (Greenwald, 1975, as cited in Dalton, 2012), meaning that some significant research-based contributions to the discipline and answers to the review questions might not have been included. Second, our search might have missed recently published articles. That is, it is not atypical for new journal publications to be delayed in appearing in databases. Once published in journals, it may take a few weeks or so for them to be visible in databases. The third risk has to do with search inputs. We might have missed studies that did not contain any of our search terms in their titles and abstracts. Fourth, only written English literature was included, meaning that relevant publications in other languages were missed.

Discussion
Language education research contributes immensely to our collective efforts to minimise the disruption to language learning brought about by the Covid-19 pandemic. Thirty-three of the 38 studies were published in 2020 and the remainder in 2021. Thirty-six studies were conducted in L2 contexts, most of which were English as a Foreign Language (EFL) and English as a Second Language (ESL) (n = 29). Other languages in focus included Arabic, Indonesian, Italian, Portuguese, and Spanish.

Research Focuses and Educational Contexts
Unlike previous trend analyses of language education research (Abdel Latif, 2018;Stapleton & Shao, 2017;Thumvichit, 2020), our analysis did not yield a wide range of research categories because we committed to the broader focuses of research rather than specialised niches. The three research focuses that emerged from open coding covered the studies well. Figure 2 shows that Learner was the most common research focus in the Covid-19 setting (n = 17). This label covered studies that investigated learners' cognition, affect, and experiences in relation to emergency language learning. Similarly, Teacher (n = 13) referred to studies that investigated teachers' cognition, affect, and experiences in relation to teaching practices during Covid-19.
Impact of instruction, a research focus adapted from Instructional effects (Stapleton & Shao, 2017), covered studies that were dedicated to determining the impact of a particular instructional method, technique, approach, tool, or material on learners' target language performance (e.g., ability, skill, knowledge, proficiency). Although this type of study has dominated language teaching research for more than a decade (Stapleton & Shao, 2017), only eight studies in this area were identified in our review. The ultimate aim of these studies was to seek language teaching solutions for emergency remote education and probably future distance education.
To be more precise, Impact of instruction studies measured learners' performance using a systematic procedure. Despite the sudden shift in teaching mode, some researchers did manage to carry out experimental research with a systematic assessment procedure (e.g., pre-and post-test, a series of exams).

Distribution by Research Focus (n = 38)
Learner 45% Teacher 34% Impact of instruction 21% Figure 3 presents the educational contexts with which the studies were concerned. Over 70% of the studies focused on the tertiary level (n = 27), covering both undergraduate and graduate levels. Although many chose LANGUAGE EDUCATION IN EMERGENCIES this educational setting on the basis of convenience, some studies considered more than one stage (n = 4). An example of this is Moser et al.'s (2020) national-scale study involving language teachers from PreK-12 and post-secondary education in the USA. This study was the only one that included the pre-primary level. Only a small number of studies focused on primary (n = 2) and secondary (n = 3) education. This may be because, in most instances, researchers are faculty members at higher education institutions, and thus doing research activities in their own contexts was more viable, especially during these troubled times. On this issue, research collaboration with other education sectors may not be as common as it was. While almost all of the studies focused on the formal education context, Motteram et al.'s (2020) study was situated in a refugee camp in northern Jordan. The researchers referred to this context as "informal school", taught by Syrian teachers recruited by NGOs; thus this context is categorised as Other. Farrell and Stanclik's (2021) study was also considered Other because detailed information about the participant's context ("a prominent English language institution") was not disclosed in an effort to keep the identity of the participant confidential.

Learning Tools in Focus
Following the pandemic, several remote learning tools were put into operation. Amin and Sudari (2020) explored Indonesian EFL learners' preferences regarding the implementation of three main types of learning platforms: video conferencing apps, learning management systems (LMSs), and mobile messenger apps. Although digital platforms are yet to fully match face-to-face classrooms, each type of digital platform has its own learning benefits. Apart from widely used tools like video conferencing apps (e.g., Google Meet, Microsoft Teams, WebEx, Zoom) and LMSs (e.g., Blackboard, Google Classroom, Moodle), mobile messaging apps can play a substantial role in facilitating language learning during lockdown. In Amin and Sudari's (2020) study, an instant messaging app like WhatsApp was preferred by students considering attention to the meaning of the language (meaning focus), opportunity for engagement with the language (learner fit), positive effects of computer-assisted language learning (CALL) activity (positive impact), and access to resources (practicality). These findings are confirmed by Ajmal et al.'s (2020) study investigating EFL learners' perceptions of LMSs and mobile messaging apps. They found that students were in favor of WhatsApp because of its practicality. While LMSs facilitate interaction and collaboration, mobile messaging apps are a great supplement to any remote learning circumstance, as they are highly accessible and do not require advanced knowledge to operate. Language learning affordances associated with social media, also discussed as social media language learning (SMLL), have recently become a topic of interest in CALL research (Reinhardt, 2020). In the context of forced remote teaching, social media has been another channel used by learners. Al-Shammari (2020) researched how students at a Kuwait international law school utilised social media to learn English. Increased use of social media to facilitate learning was reported, especially among graduate students. It was also found that students were more active when learning English through social media. In other words, WhatsApp, social media, and other standard digital platforms should not be underestimated. Most students are already familiar with their functions, making adaptation to remote learning less painful. In cases in which classes are conducted asynchronously, these tools add extra channels for communication to compensate for the loss of face-to-face interaction. Teachers should stay connected with students throughout the lesson, especially during the closure of educational institutions.
One study of pre-service teachers was selected for review. The context of teacher education differs from that of traditional language learning in that it demands field-experience learning. In a normal situation, pre-service teachers are required to teach actual classes under the guidance of their supervisor. They may be given opportunities to observe classrooms before gaining field experience first hand. However, with the suspension of face-to-face instruction, support was needed in those areas on top of conventional remote learning. Kamhi-Stein et al. (2020) implemented Mursion, a mixed-reality simulation platform, and compared the experiences of MA TESOL pre-service teachers with teaching experience and those without teaching experience. This platform is designed to support pre-service teachers in practicing their instruction through the artificial representation of a variety of classroom events, making it an especially useful tool during the closure of educational institutions. Despite its benefits in improving pre-service teachers' confidence in teaching practice (Hudson et al., 2018), some issues related to its use have been raised, such as the awkward-looking avatars (Dalinger et al., 2020), the avatars' lack of natural physical movement (Hudson et al., 2018), and different gender treatment (Black et al., 2016). Kamhi-Stein et al. (2020) found that Mursion was perceived differently by the two groups. Those with teaching experience felt that the platform did not reflect reality, and thus considered their experience inauthentic. On the other hand, those with limited or no teaching experience reported that the platform helped them to develop the confidence needed for real classroom instruction. The researchers made it clear that the current version of the platform is yet to replace a face-to-face teaching practicum, but that during lockdown, virtual simulations allow pre-service teachers to practice instruction even without actual classes.

Self-Access Language Learning
The abrupt shift in learning mode and the growing tension of the Covid-19 situation prompted a demand for intervention. Advising in language learning (ALL), as a form of intervention, is known to have many benefits, including providing psychological support (Kato & Mynard, 2016), addressing learner needs (Mynard et al., 2018), and promoting learner autonomy (Carson & Mynard, 2012). Guban-Caisido (2020) examined the implementation of a language advising program as a response to the abrupt shift to self-access language learning (SALL) at a Philippine university. The context of Guban-Caisido's (2020) study is different from previously reviewed studies, in that synchronous online classes were replaced by SALL to avoid technical issues. Students were provided with take-home packages containing the course syllabus, materials, and links to video recordings uploaded to an LMS. They were expected to stay at home and study on their own.The ALL implemented here allowed students to reflect on their learning experience and identified students' needs and difficulties. In addition to that, ALL allowed students to reflect on their own learning in order to enhance their learning experience. That is, as the advisor encouraged students to reflect on their language learning experience, students frequently realised that they needed to consider other resources as complements to SALL.
Mideros (2020) explored how Spanish L2 learners in Trinidad and Tobago used supplementary learning resources in addition to their formal synchronous classes. The findings indicated that students tended to rely heavily on materials shared by teachers, such as presentation slides, extra exercises, and links to explanations on websites. We argue that providing students with self-study materials is important, but teachers should raise students' awareness of the availability of other resources. Teachers may overlook this aspect of online teaching because it often takes place outside of their synchronous classes. Research has suggested that computerassisted instruction, if done systematically, has the potential to raise students' linguistic awareness and foster collaboration with other students, leading to the overall enhancement of learner autonomy (Benson, 2013).

Social Support
Given the difficulties and limitations in emergency language learning, learner engagement has been regarded as an important element in improving academic achievement in the target language (Luan et al., 2020). Engagement is substantially influenced by learners' interactions with their learning environment (Reschly et al., 2020). This link is receiving increasing attention from scholars, as its association with EFL learners' perceived social support creates space for research on the abrupt transition to remote learning. Luan et al. (2020) explored the interplay between learners' perceived social support and learner engagement in the LANGUAGE EDUCATION IN EMERGENCIES Chinese context.The findings showed that students who received support from their teachers and peers demonstrated higher engagement in online language classes. It is important to note that the learner engagement mentioned here largely refers to learning activities, including involvement in instructional content, and interpersonal exchange with teachers and peers. As the link between learner engagement and teacher support has already been established, such findings were further elaborated by Hew's (2016) study suggesting that learner engagement depends on the extent to which teachers are willing to communicate with students, their teaching enthusiasm, and the learning resources that they provide to facilitate remote learning. During the time of Covid-19, students as well as teachers need more support than before, and one thing that teachers can do is to reinforce positive engagement by guiding students through meaningful and motivational activities (Soffer & Cohen, 2019) or adding elements of interaction to the syllabus (Veletsianos et al., 2015).

Learning Styles and Emergency Language Learning
The relationship between learners' preferred learning styles and remote learning has also been a subject of study. To provide an evaluation of an online ESL course, Syahrin and Salih (2020) investigated whether the course content, activities, and functions of an LMS responded to students' learning styles. In this study, students' learning styles were identified first using Kolb's (1984) Learning Style Inventory. Since most participants were Convergent learners -learners who "draw from the learning modes of abstract conceptualisation and active experimentation" (Richmond & Cummings, 2005, p. 48) -online language lessons focusing on lectures or receptive skills (reading and listening) might not disrupt their language learning as much as one may expect. This work sparks our curiosity about which learning style is most suitable to the context of emergency language learning. Based on Richmond and Cummings's (2005) summary of Kolb's (1984) learning styles in online education, Assimilative learners seem most resilient during emergency language learning, as they tend to rely on abstract concepts and ideas rather than interaction.

Contributions to the Local Community
Zheng (2020) examines the multilingual translation efforts made by a group of foreign language university students. This study is one of a small number that address how language education contributes to the local community in times of crisis. It was situated in Shanghai, where multilingual public services are very much needed. In this study, student volunteers provided translation support in ten languages when communities had to deliver important messages regarding disease control and precautionary measures to foreign residents and visitors. Since the translation team was assembled hastily, it comprised a number of students, many of whom had limited experience with the target language. Therefore, they ran into some linguistic difficulties including accents, an inability to discuss the specificities of everyday life in the target language, and naturalness in translation. To overcome these difficulties, they sought help from competent users of the target language. Similar to emergency language teaching, multilingual emergency translation is characterised by a tension between rising demands for translators and inadequate resources. This study suggests that even in times of crisis, there is room for language education to contribute to local communities.

Dealing with Changes
Farrell and Stanclik (2021) report on a case study of an early-career EFL teacher at an English language institution in North America in terms of philosophy, principles, theory, practice, and the space beyond practice. The participating teacher's reflections on each of these five elements highlighted the connections between his stated principles and actual practices. Here we will focus our discussion on those practices shaped by his principles and the Covid-19 pandemic. Despite the abrupt transition to online teaching, he retained almost all of his principles and put them into practice. They were observed through the following practices: • Keeping to lesson plans • Providing feedback • Correcting errors • Incorporating students' cultural knowledge into activities • Clearly stating instructions • Making himself available for students • Engaging in informal interaction with students • Making sure students were still on task • Featuring group activities These practices are worth sharing because they can be implemented even in emergency teaching. Teachers may not have to compromise their principles in order to deliver online instruction effectively. The challenges posed by unprecedented crises can be embraced as opportunities to (re)discover oneself as a language teacher. By encouraging language teachers to reflect on their practice, an evidence-based practice study like this one can greatly benefit its participants and other teachers, in that they can use the information revealed to help themselves make pedagogical decisions "rather than following hunches not based on any concrete evidence" (Farrell & Stanclik, 2021, p. 12).
Insights into the nature of the abrupt transition to remote teaching can shed light on current and future practices. The transition brought several changes that language teachers needed to confront. Moser et al. (2020) launched a national survey to better understand changes in practices and perceptions of PreK-12 and postsecondary language teachers from 45 US states. As one may expect, the vast majority of participants had no online teaching experience, and consequently had to put in more time, work, and effort to deal with the changes than those with prior experience even though training sessions on using digital platforms for teaching were provided. Although teachers with prior experience did not have to make substantial changes in course design and adjustment, lack of significant differences in practice indicated that all teachers were new to emergency remote teaching. Despite embracing effective principles of remote language teaching, teachers reported that their students' academic outcomes were lower. The reason is that much of their attention was devoted to students' basic needs, meaning that teachers played a role in supporting the non-academic side of students' lives during the pandemic. The findings suggested that PreK-12 teachers were more vulnerable than postsecondary teachers and needed serious training and support. The researchers explained that PreK-12 teachers faced equity issues that were not typical in post-secondary education, including lack of technology and unsupervised students without family support. Therefore, they had to take on usual tasks such as delivering paperwork packets and meeting with individual students. Moser et al. (2020) concluded that all teachers, especially PreK-12 teachers, needed more support to enact remote teaching and achieve desirable outcomes.
Sepulveda-Escobar and Morrison (2020) explored the challenges and opportunities of virtual teaching experienced by EFL teacher candidates in Chile. Despite the school closure, some in-field teacher training programs continued. Participating student teachers sent to different schools found themselves in a variety of situations, as influenced by schools' policies. The findings indicated that the most challenging aspect of emergency language teaching was the absence of interaction with students, which may have hindered their professional development (Flores & Gago, 2020). This study uncovered two important components that could not be developed: social development (e.g., collaboration with colleagues and mentors) and personal development (e.g., managing feelings about being a teacher) (Bell, 1994). However, student teachers were given an opportunity to learn to use different digital platforms and to design their own teaching strategies to engage their students without seeing them in person.

Factors Influencing Technology Integration
It is inarguable that the success of emergency remote education has relied considerably on teachers' integration of technology into instruction, which can be influenced by various factors. Cheung (2021) conducted a case study of a secondary school ESL teacher in Hong Kong who taught synchronously via a video conferencing app, with the view to identifying the factors influencing technology integration. The abrupt transition to synchronous online learning environments resulted in fewer opportunities to interact with students and monitor their understanding. It was also found that pedagogical beliefs, the context, and professional development were largely mediated by the pedagogical use of technology. For example, an implicit belief in the grammar-translation method as the best way to help students pass the form-focused exam is likely to remain in online teaching mode. This could prevent teachers from exploring the other useful functions of video conferencing apps, such as Breakout Rooms in Zoom and Assignments in Microsoft Teams.The level of technology integration could also be influenced by the way in which educational institutions react to the call for the transition to online teaching. For example, the participating teacher in Cheung's (2021) study noted that her school was passively reacting to the call and was not seriously interested in developing an online teaching community over the long term. This indirectly affected her technology integration, despite the fact that she had all of the necessary resources. Training courses play a crucial role in elevating the level of technology integration, but only if they offer hands-on experience and if teachers are not overly busy with other workloads.

LANGUAGE EDUCATION IN EMERGENCIES
Carvalho (2020) addressed factors surrounding the infrequent use of digital resources in teaching Portuguese as a Non-Native Language (PNNL). The findings showed that attending Information Communication Technology (ICT) training sessions did not encourage the use of digital resources for PNNL teaching purposes.
Although age was not found to be a factor justifying the infrequent use of technology, younger teachers tended to be more confident in using technology than those who had been in the profession longer. Fuad et al. (2020) added that many senior teachers found it difficult to integrate digital technologies into language teaching because of their lack of previous exposure to ICT. Confidence plays a vital role in promoting the integration of technologies, and teachers' confidence can be built through training sessions that focus on technology and education. Training should be organised in such a way that it helps teachers to recognise the intellectual benefits of digital resources beyond this temporary necessity. Motteram et al. (2020) explored the challenges and possibilities in using WhatsApp to support language teacher development in a Zataari refugee camp in Jordan. Syrian teachers were recruited to teach at a support school that aimed to help relocated Syrian children catch up with the Jordanian education system and continue onto higher education. In this study, the chat history of a group of English teachers was analysed to monitor any teacher development activities carried out through WhatsApp. The findings showed that a great number of messages were concerned with language development, meaning that teachers' language knowledge and skills were being developed through the app. For example, one of the teachers expanded his linguistic knowledge by asking his fellow teachers about the word "yeah". Teachers also reflected on their training sessions, discussing cultural issues and sharing pedagogical strategies. This does not, however, suggest that teacher development should be completely online from now on, but it does indicate that an everyday app like WhatsApp can provide a great deal of support. Whether or not there are disruptions in the future, teacher development should take advantage of the support provided by digital tools.

Stress and Coping Strategies
For language teachers, few days go by without some kind of stress or difficulty (MacIntyre et al., 2019). L2 teachers face the emotional challenges of L2 teaching, such as doubting their own L2 ability, managing the diverse proficiency levels of learners, and keeping up with demanding pedagogical methods (Gkonou et al., 2020;Gkonou & Miller, 2017). The pandemic brought them even more stress, making their work and life even more difficult to navigate. MacIntyre et al. (2020) examined stress as a result of the pandemic, along with the coping strategies of language teachers from various countries. The findings indicated that workload was the most common cause of stress, followed by the health of family members. The most used coping strategy was acceptance -acknowledging the reality and trying to live with it -followed by advance planning, both of which were considered approach strategies. It is encouraging to learn that avoidant strategies like disengagement, substance abuse, and denial were least common, because these strategies often lead to undesirable outcomes (e.g., anxiety, anger, loneliness, sadness). Approach strategies, on the other hand, are likely to produce desirable outcomes (e.g., happiness, health, resilience, wellbeing). Where possible, language teachers should resist avoidant coping mechanisms and be more realistic and optimistic. Since it is not known how much longer the "new teaching normal" will last or whether the "old normal" will even return, teachers need to embrace the changes that the pandemic has introduced into education instead of fleeing from them and waiting for everything to return to normal.

Conclusion
This systematic review brings together the available language education research within the context of Covid-19 to gain a thorough understanding of the state of affairs, as well as the situation's inherent challenges and opportunities for language learners and teachers. Since the outbreak, the research community has been active in contributing to the collective effort to find solutions for the current emergency education demand. Since much of the research was undertaken in the tertiary setting, we would like to call for more research on other educational stages, especially PreK-12, whose teachers and learners are more prone to issues such as inequity. The sudden, disruptive transition to remote education represents fresh challenges and opportunities to (re)discover and (re)invent teaching and learning. The digital tools that have been implemented in emergencies like this are capable of engaging language learners in an interactive learning experience, though they are yet to replace the face-to-face environment. Technology is, however, not the only determinant of academic achievement. SALL, ALL, and other forms of outside-of-class support should be considered supplements to regular remote classes and used to stay connected with learners and provide timely support. In addition to maintaining access to learning, language education can take on the important task of providing multilingual services to aid Covid-19 prevention for foreigners. As the severity of the situation varies from case to case, there is plenty of room for research and innovation that will create novel alternatives and solutions for emergency language learning.
Language teachers may have been forced into the transition without adequate preparation, but they do not need to abandon their beliefs about what effective teaching should be. Educators should hold on to their pedagogical principles and try to put them into practice regardless of teaching delivery mode. Teacher development should also continue. Everyday mobile messaging apps can serve as a springboard to the more versatile learning tools. With their wellbeing as paramount, teachers should accept the changes and take them as opportunities to learn and explore new possibilities. Although it is hoped that the current pandemic will end as soon as possible and a crisis of this kind will never happen again, remote teaching should no longer be considered a temporary practice. Educators should take the lead to establish a community of practice for remote teaching as it can, in fact, be a viable solution for many normal educational scenarios. In closing, we acknowledge and appreciate every contribution that has been made to addressing the disruption and maintaining access to learning, including those that are not mentioned in this review.