A decade of
innovation in language education: Insights from eight leading CALL-related
journals
Junjie Gavin Wu1, Sangmin-Michelle
Lee2, Di Zou3, and Junhua Xian1*
|
*Correspondence: junhua.ax@gmail.com |
|
1Macao
Polytechnic University, Macau SAR 2Kyung
Hee University, South Korea 3The
Hong Kong Polytechnic University, Hong Kong SAR |
Abstract
Technological
advancements—from mobile and extended reality (XR) to generative artificial
intelligence (GenAI)—are central to innovation in computer-assisted language
learning (CALL). This systematic review conducted a thorough analysis of 1,098
papers from eight leading CALL-related journals (2015 – 2025). With BERTopic
modeling, core research topics and their evolution were identified. The study
uncovered nine key topics across the eight journals, such as affect in L2
speaking, teacher development, mobile and immersive technologies, and more.
Findings suggested that different topics undergo different developmental life
cycles and GenAI has emerged as a unique transformer of innovation in CALL. The
review concludes with practical implications for both teachers and researchers.
Keywords:
innovation in CALL, topic modeling, review, generative AI, core research
topics
But the good thing is, with innovation, there isn’t a last nugget.
– Jeff
Bezos, Founder of Amazon
1.
Introduction
With the emergence and
swift adoption of generative artificial intelligence (GenAI), the field of
computer-assisted language learning (CALL) has entered a golden age of
research. In 2018, Colpaert argued that “the CALL field remains vulnerable to
absorption by other disciplines due to its feet of clay” (p. 483), yet this
perspective may no longer accurately reflect the dynamics of contemporary
applied linguistics (AL). Instead, we contend that other AL subfields are
actively advancing CALL by sharing, integrating, and co-constructing theories,
findings, pedagogical practices, and interdisciplinary insights.
A review of the latest
Web of Science (WoS) report (Table 1) reveals that most of the top 10 journals
in education and linguistics are technology-oriented, underscoring this shift.
Interestingly, nearly all top 10 journals in educational studies are indexed in
SSCI (Social Sciences Citation Index), with the exception of Smart Learning Environments, which is
indexed in the ESCI (Emerging Sources Citation Index). With the aims and scope
on the intersection of education and technology, its appearance on the list
provides evidence of the growing academic value of technology in education.
Among linguistics journals, in addition to CALL-specific publications such as Computer Assisted Language Learning and ReCALL, others such as RELC Journal in Table 1 also place a
strong emphasis on the use of technology in language learning and teaching.
Moreover,
an examination of the more comprehensive Scopus database and its latest
CiteScore metrics for educational journals shows that certain Scopus-indexed,
non-SSCI publications rank at the forefront. A prime example is Computers and Education: Artificial
Intelligence, which holds the number 2 position among 1,957 journals with a
CiteScore of 28.7 as of November 2025 (noting that these scores are updated
every few months). Taken together, these trends suggest that technology has
permeated nearly every facet of educational research, especially language
education.
Table 1.
Top 10 educational and linguistics journals in the WoS database
|
No. |
Educational Journal |
Index |
2024 Impact Factor |
|
1 |
International Journal of
Educational Technology in Higher Education |
SSCI |
16.7 |
|
2 |
ESCI |
12.1 |
|
|
3 |
Educational Psychologist |
SSCI |
11.4 |
|
4 |
SSCI |
10.6 |
|
|
5 |
SSCI |
10.5 |
|
|
6 |
SSCI |
9.9 |
|
|
7 |
British Journal of Educational Technology |
SSCI |
8.1 |
|
8 |
International Journal of STEM Education |
SSCI |
8.0 |
|
9 |
Review of Educational Research |
SSCI |
7.4 |
|
10 |
International Journal of Management Education |
SSCI |
7.4 |
|
No. |
Linguistics Journal |
Index |
2024 Impact Factor |
|
1 |
Transaction on the Association for Computational Linguistics |
SSCI |
6.9 |
|
2 |
Computer Assisted Language Learning |
SSCI |
6.6 |
|
3 |
RELC Journal |
SSCI |
6.2 |
|
4 |
ReCALL |
SSCI |
5.7 |
|
5 |
System |
SSCI |
5.6 |
|
6 |
Assessing Writing |
SSCI |
5.5 |
|
7 |
Studies in Second Language Learning and Teaching |
SSCI |
5.4 |
|
8 |
Computational Linguistics |
SSCI |
5.3 |
|
9 |
Language Teaching |
SSCI |
5.1 |
|
10 |
Studies in Second Language Acquisition |
SSCI |
4.9 |
SSCI=Social Sciences
Citation Index, ESCI=Emerging Sources Citation Index
Over the past three decades, technological
advancements have introduced unexpected, profound innovations to language
education. From early CD-ROMs and Language Massive Open Online Courses (LMOOCs)
to mobile learning tools such as Padlet, and now to the integration of GenAI
and extended reality (XR), language education has progressed towards more
interactive, immersive, adaptive, and authentic modes of learning, compared to
the traditional, decontextualized rote learning (Zhang & Zou, 2022). These
innovative technologies such as GenAI and XR are innovating teaching and
learning both inside and outside the classroom, influencing every aspect – from
lesson preparation and in-class activities to after-class assignments (Wu et
al., 2024). Recent studies have provided empirical evidence of their unique,
innovative affordances, such as delivering customized, real-time feedback,
boosting learning motivation, increasing language proficiencies, and improving
student interactions (e.g., Kohnke et al., 2023; Wu et al., 2023; Zhang et al.,
2024); however, challenges have also been documented including shortages of
equipment, insufficient knowledge, and a lack of expertise for effectively
planning, integrating, and implementing these technologies across diverse
educational contexts (e.g., Lee & Wu, 2024; Wu & Miller, 2025).
With these technologies, innovation is now a
central theme in CALL and in contemporary language education. Over the past
decade, Reinders and his colleagues have collected a series of reports on
innovative teaching and learning practices across various contexts such as the United
Kingdom (Chong & Reinders, 2024), Korea (Reinders et al., 2025), and
Vietnam and Cambodia (Phung et al., 2024). In these edited volumes, technology
consistently serves as a crucial driver of innovation, empowering both learners
and teachers with novel pedagogical approaches. Yet, as Carless observes,
“[i]nnovation is extremely difficult to engineer successfully” (p. 1),
underscoring the challenges in implementing innovative practices.
Complicating the issues is the absence of a
comprehensive, systematic synthesis of innovative, technology-empowered
practices in language education. Existing studies remain fragmented since they
are often limited to case studies or anecdotal reports, without offering a
holistic understanding of core themes, their interconnections, and temporal
evolution.
To fill this gap, the present review study offers
a rigorous, data-driven overview of the existing innovation-related studies in
CALL. Specifically, a corpus of articles from eight leading CALL journals is selected.
Furthermore, advanced natural language processing (NLP) techniques are applied
to shed light on the core research topics and their relationship and
developments. The following three questions guided the research:
RQ1:
What are the core research topics of innovation in language education?
RQ2:
What structural relationships exist among these core topics?
RQ3:
How have the core research topics evolved?
2. Dataset and methods
2.1
Research Design
This study employed a
systematic literature review method to map the conceptual landscape and
research trends concerning innovation in CALL. The entire process of
identification, screening, eligibility assessment, and inclusion of relevant
studies was conducted in strict accordance with the Preferred Reporting Items
for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Following the
systematic literature retrieval and screening, the study incorporated a
computational thematic analysis using the bidirectional encoder representations
from transformers for the topic modeling technique (BERTopic). This integration
of systematic review principles with NLP was designed to objectively identify,
quantify, and interpret the latent thematic structure within the corpus of
selected literature (Devlin et al., 2019), thereby providing a robust and
data-driven overview of the research domain (Xian et al., 2025). The overview
of the entire research method is shown in Figure 1.
Figure
1. Overview of the research method
2.2
Data Source and Search Strategy
The literature search
was performed exclusively within the Scopus database, renowned for its
extensive coverage of high-quality, peer-reviewed literature. The search
strategy was designed to capture a focused yet substantial body of research
from the core journals in the field of CALL. The search was confined to the
eight representative CALL-related journals: Innovation
in Language Learning and Teaching, Computer
Assisted Language Learning, Language
Learning & Technology, ReCALL,
International Journal of
Computer-Assisted Language Learning and Teaching, CALL-EJ, Technology in
Language Teaching and Learning, and The
JALT CALL Journal.
The
search query utilized a combination of key terms related to the core concept of
“innovation” (innovation, innovative, innovative learning, innovative teaching)
searched within the article “titles, abstracts, and keywords”. Several filters
were applied to refine the results: Publication Window: January 1, 2015, to
October 1, 2025. Document Type: “Article” to ensure the inclusion of original
research. Language: “English”.
2.3
Screening and Eligibility Criteria
The study selection
process involved a two-phase screening against pre-defined eligibility criteria
to identify the most relevant studies.
2.3.1
Inclusion Criteria
·
The review includes
empirical studies (qualitative, quantitative, or mixed methods) and conceptual
papers that centrally address the theme of “innovation” or “innovative”
practice in language education.
·
Studies must focus on
the application of technological tools or platforms in language learning or
teaching contexts.
·
Publications must fall
within the specified date range, meet the language requirements, and conform to
the predetermined document types.
2.3.2
Exclusion Criteria
·
Publications such as
editorials, book reviews, or letters are excluded.
·
Articles where the
terms “innovation” or “innovative” are only peripherally mentioned without
constituting a substantive focus of the study are eliminated.
·
Studies whose primary
focus extends beyond language learning or teaching are not considered for
inclusion.
2.4
Study Selection Process
The study selection
process, detailed in the PRISMA flow diagram (Figure 2), was rigorously
managed. In addition, the number of articles selected from each of these eight
journals is as follows: Innovation in
Language Learning and Teaching (n = 460), Computer Assisted Language Learning (n = 243), Language Learning & Technology (n = 67), ReCALL (n = 60), International
Journal of Computer-Assisted Language Learning and Teaching (n = 59), CALL-EJ (n = 118), Technology in Language Teaching and Learning (n = 27), The JALT
CALL Journal (n = 64).
Figure
2. PRISMA flow diagram of the study selection process
2.5 Data Extraction and Analysis
2.5.1
Data Extraction and Corpus Formation
For the final set of 1,098
included studies, a structured data extraction was performed to compile the
corpus for thematic modeling. The key textual components of each article were
extracted and consolidated into a single composite document per article. This
approach capitalizes on the fact that these elements concisely encapsulate the
study’s core content and themes, making them ideal for topic modeling. This
collection of documents constituted the primary corpus for subsequent
computational analysis. Relevant metadata, such as publication year and source
journal, were also extracted to facilitate the interpretation of the resulting
topics over time and across publication venues.
2.5.2
Text Pre-processing
To prepare the textual
corpus for subsequent analysis, a series of standard natural language
pre-processing steps was implemented to refine the data, reduce noise, and
focus on meaningful semantic content (Gardazi et al., 2025). This procedure was
conducted using the Python programming environment and involved several key
stages. Initially, all text was converted to lowercase to ensure consistency.
This was followed by tokenization, which segmented the continuous text into
individual word units. Subsequently, we removed common English stop words
(e.g., “the,” “and”) as well as frequently occurring yet semantically weak
vocabulary specific to the research domain (e.g., “study,” “paper,” “findings”)
to accentuate words of substantive meaning. Finally, the words underwent
lemmatization, a process that reduces them to their base or dictionary form
(for instance, converting “innovative” and “innovation” to “innovate”), thereby
consolidating different morphological variants at a semantic level and
effectively refining conceptual expression. The clean and standardized corpus
resulting from this pipeline provided a reliable foundation for the in-depth,
semantic-based topic modeling that followed.
2.5.3
Topic Modeling with BERTopic
Thematic analysis was
conducted using BERTopic, an advanced topic modeling technique that leverages
contextual semantic understanding. Unlike traditional frequency-based models,
this approach captures the deep semantic relationships between words, resulting
in more coherent and interpretable topics. Our modeling process encompassed
several key phases. First, we utilized the all-MiniLM-L6-v2 model from the
sentence-transformers library to convert the pre-processed text corpus into
high-dimensional document embeddings. This step represented the semantic content
of each article as a dense numerical vector, laying the groundwork for
subsequent clustering. Next, we employed the uniform manifold approximation and
projection (UMAP) algorithm to reduce the dimensionality of these
high-dimensional vectors (Raman et al., 2024). This process aimed to compress the data into a
lower-dimensional space more suitable for clustering while preserving the most
significant semantic structures between documents. Subsequently, the
hierarchical density-based spatial clustering of applications with noise
(HDBSCAN) algorithm, a density-based clustering method, was applied to the
reduced data to identify topic clusters (Raman et al., 2024). A key advantage of
this algorithm is its ability to automatically identify dense clusters of
arbitrary shapes and to classify documents that do not firmly belong to any
strong theme as outliers, thereby enhancing the internal purity of the final
topics. Finally, for each cluster identified by HDBSCAN, a class-based term frequency-inverse document frequency (TF-IDF)
technique was employed to extract the most representative keywords (Raman et
al., 2024). This method
evaluates the importance of words within a specific cluster relative to their
frequency across the entire corpus, generating a set of highly discriminative
and representative labels for each topic, which facilitates interpretation and
analysis by the researchers.
3. Results
3.1 Determining
Core Topics
Table 2 reveals the nine core research topics identified
through BERTopic analysis, along with their basic information and
characteristics. The core terms of each topic in Figure 3 and their c-TF-IDF scores reveal distinct boundaries
between topics. Below is a detailed presentation of these five topics:
Topic 0 (T0) Affect in L2 speaking
Topic
0 represents the most extensively researched area within innovative CALL
studies. As evidenced by the keywords in Table 2, this topic is characterized
by a strong emphasis on psychological constructs, notably “self,” “anxiety,”
and “motivation” in the context of L2 speaking development. The appearance of
“teacher” among the key terms further underscores the educator’s role in
mediating these affective factors during speaking practice. Collectively, these
lexical patterns reflect a student-centered pedagogical approach that
prioritizes learners’ emotional and self-regulatory experiences, aligning with
a broader shift in innovative CALL approaches toward affective and
psychologically informed language teaching (e.g., Fathi et al., 2024; Zhang et
al., 2024).
Topic 1 (T1) Teacher development & CALL
pedagogy
This
research topic emphasizes teacher professional development and pedagogical
innovation in CALL. Keywords such as “finding” and “data” highlight the
empirical nature of this topic, and teacher education has its foundational role
in the innovative learning and teaching: before technologies can be implemented
into learning, teachers must first be equipped with the necessary cognitive,
pedagogical, and technological competencies (Lee & Wu, 2024; Meihami, &
Esfandiari, 2021). Frameworks such as TPACK (Technological Pedagogical Content
Knowledge) and SAMR (Substitution, Augmentation, Modification, Redefinition)
have been widely explored and revised in different technological environments,
from computer technology to GenAI and XR (Lee & Wu, 2023). This strand
highlights the crucial role of teachers as agents of innovation and
conceptualizes innovation as a process mediated by teacher readiness. This body
of work investigates how teachers should critically evaluate digital tools and
adapt them to diverse learning contexts (Choi et al., 2025).
Topic 2 (T2) Mobile and immersive technologies
with vocabulary learning
This
topic is related to new technologies, especially virtual reality (VR) and
augmented reality (AR), reflecting a tool-oriented research agenda within
innovation studies. Together with “vocabulary”, these keywords showcase that
these tools are used to create immersive experiences for learners to develop
their vocabulary knowledge and skills (Lee et al., 2024). Recognized for the
capacity to simulate authentic, low-stakes scenarios in the target language, VR
and AR empower learners to engage in contextualized practice (Song et al.,
2023). Thus, innovation here is realized through the lens of specific digital
tools and the interactive environment, positioning technological mediation as
central to the learning process.
Topic 3 (T3) Intercultural communicative
competence through telecollaboration
The
keywords “intercultural,” “telecollaboration,” “Chinese,”, and “English”
delineate a topic that centers on cross-cultural language learning where
technology plays an important role. Over the past decade, we have seen a wide
array of telecollaboration research using different digital tools from the
traditional computer-mediated communication to mobile-assisted interactions,
and more recently VR environments (Wu et al., 2023; You et al., 2025). Within these innovative
learning configurations, learners usually participate in synchronous,
multimodal communication with partners from other cultural contexts. Learners
develop not only linguistic skills, but also the awareness of global
citizenship and other soft skills needed to thrive in the global job market.
Topic 4 (T4) Course design for online, blended, or
flipped learning
This
topic of CALL research addresses innovation at a curricular level. It explores
the optimal design and implementation of online courses, blended learning, and
flipped learning (e.g., Amiryousefi, 2019; Chu et al., 2025). In particular,
LMOOCs represent an innovative approach to achieving a large-scale mode of
delivery (Wright & Furneaux, 2021), contributing to the democratization of
education worldwide.
Topic 5 (T5) Collaborative writing and automated
writing evaluation
This
topic explores the integration of human-centered (“peer” and “collaborative”)
and technology-driven (automated writing evaluation, AWE) methods to innovate
writing practices. Previous CALL research has been characterized by different
technological platforms such as Google Docs, Wiki, and online blogs (Lai, 2023;
Li & Zhu, 2017), while AWE has also captured attention for its capabilities
to provide feedback on low-level writing skills such as grammar (Link et al.,
2022). The innovation dimension of this research lies in its focus on how
learners can co-construct meaning and writing pieces with the support of
technology.
Topic 6 (T6) GenAI in Language Education
This
line of innovative CALL research focuses on the use of GenAI to provide
personalized learning. The keyword profile for this topic is dominated by
specific system names such as “ChatGPT” and “AI”, so it suggests that the field
is in a premature and exploratory phase of research. This emphasis on tools
suggests that attention remains on understanding the capabilities and
affordances of the technology (e.g., Zeevy-Solovey, 2024; Zhou & Du Preez,
2025). This is further corroborated by the semantic similarity analysis in
Figure 4, which reveals the conceptual isolation of GenAI from other topics,
which will be discussed in the subsequent sections.
Topic 7 (T7) Video captions for EFL input
comprehension
This
topic examines the use of multimodal input, particularly captioned videos, to
enhance receptive skills, including listening and reading. Scholars such as Teng
(2025) and Lin (2022) have conducted studies to compare how different types of
captions provide audiovisual scaffolding for language learning. Key words like
“test” and “comprehension” further point to an outcome-oriented approach to
innovation that focuses on the measurement of learner gains in the technology
environments.
Topic 8 (T8) Digital game-based learning and
creative project-based language learning
The
keywords under this topic in Table 2 signal the creation of experiential and
expressive language learning experiences in CALL studies. Words such as
“creative,” “game,” “enjoyment,” “story,” and “narrative” reflect pedagogical
approaches such as digital game-based learning, project-based learning, and
digital storytelling learning, which emphasize agency, motivation, and
multimodal literacy (Li et al., 2022; Lee, 2019; Tour et al., 2021).
Innovation here is manifested as the design of
personally interesting and emotionally engaging learning experiences, where
technology functions as a canvas for creation. Also, gaming elements such as
reward systems, real-time interactions, and scaffolded challenges have played a
role in fostering authentic learning opportunities for learners to develop
language and creativity skills (Foroutan Far & Taghizadeh, 2024).
Table
2. Nine core topics of innovation in eight CALL journals
|
Topics |
Document Distribution |
Topic Name |
Key Terms |
|
0 |
193 |
Affect
in L2 speaking |
L2, speaking,
efl, English, self, teacher, anxiety, motivation, foreign, participant |
|
1 |
163 |
Teacher
development & CALL pedagogy |
Teacher,
English, teaching, call, practice, classroom, development, finding, data,
approach |
|
2 |
161 |
Mobile
and immersive technologies with vocabulary learning |
Technology,
vocabulary, mobile, VR, digital, teacher, AR, based, English, game |
|
3 |
105 |
Intercultural
communicative competence through telecollaboration |
Chinese,
intercultural, English, teacher, telecollaboration, data, CLIL, analysis,
dynamic, participant |
|
4 |
92 |
Course
design for online, blended, or flipped learning |
Online,
teacher, lmooc, course, teaching, flipped, blended, English, design, EFL |
|
5 |
78 |
Collaborative
writing and automated writing evaluation |
Writing,
feedback, peer, EFL, collaborative, dmc, awe, text, performance, analysis |
|
6 |
64 |
GenAI
in Language Education |
AI,
chatgpt, genai, tool, artificial, writing, intelligence, teacher, generative,
education |
|
7 |
43 |
Video
captions for EFL input comprehension |
Reading,
listening, comprehension, EFL, caption, video, strategy, vocabulary, test,
participant |
|
8 |
21 |
Digital
game-based learning and creative project-based language learning |
Creativity,
creative, project, digital, game, enjoyment, approach, story, narrative |
Figure
3. The core keywords for each topic and their c-TF-IDF scores
3.2 Topic
Similarity Analysis
The similarity matrix
heatmap (Figure 4) reveals a clear hierarchical structure among the nine
topics, featured by clusters of strong associations alongside more isolated
topics. The overall matrix exhibits a block-like distribution pattern,
suggesting the presence of several conceptually cohesive subsets with varying
degrees of inter-cluster relatedness. The main diagonal, representing each
node’s self-correlation, maintains theoretical perfect correlation as expected.
The nodes T0 and T1, and T1 and T3, demonstrate the highest degree of
similarity (score of 1.00), with a correlation coefficient exceeding 0.92,
forming the most strongly correlated pair. Closely following are pairs such as
T1 and T4, T2 and T3, T0 and T7, T0 and T3, and T3 and T4 with similarity
values ranging between 0.90 and 0.92, suggesting a tightly knit core cluster.
At the moderate association level, connections such as T0 and T2, T4 and T5,
and T3 and T8 show a correlation of approximately 0.85, while others such as T1
and T5, T2 and T7, and T3 and T6 fall within the 0.8 to 0.85 range, reflecting
looser but still notable linkages. In contrast, T6 (GenAI in Language
Education) displays generally weaker connections to the other nodes, with
similarity coefficients mostly below 0.8, appearing as a distinctly
lighter-shaded region in the color gradient and highlighting its relative
conceptual independence.
Figure
4. Heatmaps of semantic similarity between topics
3.3 Topic
Evolution Over Time
Figure 5 showcases a clear
technology-driven development of innovations in language education from 2015 to
2025. Specifically, T6 demonstrates an explosive growth pattern, with minimal
scholarly attention prior to 2023, followed by a rapid rise as a dominant
research theme, a trajectory that closely aligns with the widespread adoption
of GenAI technologies. By comparison, T0 and T2 maintain consistently high
research interest throughout the period, indicating their transition into a
relatively mature phase of development. Notably, T1 and T4 exhibit synchronous
growth between 2020-2022, driven by the shift to remote learning during the
pandemic.
In
contrast, T8 shows a steady, linear growth trajectory, suggesting sustained and
stable development in this research area. Furthermore, T3 displays a gradual
upward curve, reflecting its progressive accumulation of academic attention as
an emerging research direction. However, T5 and T7 maintain relatively stable
fluctuations, indicating these traditional research fields have entered a developmental
plateau. In summary, the data clearly reveal a dynamic balance between emerging
and established themes: while technology-intensive topics accelerate rapidly,
foundational pedagogical themes continue to evolve steadily, together shaping
the advancing frontier of language education.
Figure
5. Topic evolution over time
4. Discussion
4.1 Thematic Structure and Trajectories of Innovation
in CALL
Evidenced
in Figure 4, the review discovers that T1 (Teacher development & CALL
pedagogy) is a central hub in the landscape of innovation within the eight CALL-related
journals. Together with T0 (Affect in L2 speaking), T3 (Intercultural
communicative competence through telecollaboration), and T4 (Course design for
online, blended, or flipped learning), these topics form a tight conceptual
cluster. This clustering suggests that over the past decade, the CALL journals
prioritized the augmentation of L2 pedagogical practices, instead of centering
on the technological tools per se.
Several trends within this cluster should be highlighted. First, teacher
education and learner psychology have remained prominent and seem that they
will continue to receive attention in innovative CALL studies (Figure 5).
Second, telecollaboration has sustained its popularity for language learners to
practice intercultural and intracultural communicative competence (e.g., Priego
& Liaw, 2017; Wu et al., 2022). It is worth noting that Figure 5 shows an
increasing trend from 2022. One possible reason is the growing availability of
AR and VR that overcome the limitations of traditional video-based chatting by
extending learning opportunities to enable learners to perform actions,
experience virtual scenes, and interact with virtual elements or characters
(Wu, 2021; Wu & Lee, 2025). Third, MOOCs, flipped learning, and blended
learning were widely explored due to the normalization of computer and mobile
technologies. However, as Figure 5 shows, this line of research has plateaued
compared to other topics. Perhaps this is related to the fact that many
existing studies, not just language research, have demonstrated that these
three pedagogical methods are not as innovative nor effective as recent
emerging methods (Hew & Cheung, 2014).
The keywords in Table 2 and Figure 3 further
suggest a spectrum of research maturity
for the nine topics. T2 and T6 with emerging technologies are marked by
tool-oriented terms such as VR, AR, and AI. In contrast, T5
(Collaborative writing and automated writing evaluation), T7 (Video captions
for EFL input comprehension), and T8 (Digital game-based learning and creative
project-based language learning) from Figure 4 represent more established
streams. They also have weaker links with emerging technologies. For T5,
automatic writing tools such as Grammarly emphasize the correction of errors
while overlooking the importance of ideas and content co-construction (Dizon
& Gayed, 2021). Similarly, video captions and digital game-based learning
are largely limited to computers (Dashtestani, 2022). Only recent researchers have
begun discussing the potential of aligning these conventional research areas
with innovative technologies (e.g., Wu et al., 2025).
In terms of T2 and T6, there are two distinctive
modes of innovation. T2 exemplified dependent innovation, which has strong
associations with various topics in Figure 4. This highlights that mobile and
immersive technologies are being used to enrich existing learning contexts for
learning practices (Hwang & Lee, 2024; Wu & Miller, 2021). On the
contrary, T6 demonstrates independent innovation, pronounced by its weak
semantic connections with other topics. This isolation points out that GenAI is
not being treated merely as an extension of existing CALL technologies. Rather,
it functions as a unique and potentially disruptive force to innovating
language learning paradigms (Huang et al., 2023).
A few factors have contributed to this
independence. First, differing from traditional automated writing tools that
provide corrections on linguistic errors, GenAI has the great potential to
engage learners by generating and interacting with the learner in a dialogic
and natural manner. This offers brand new opportunities for moving away from
tool-based learning towards learner-technology collaboration (Chen et al.,
2022). Second, as mentioned, the established domains are so far largely refined
to computer and mobile technologies, but GenAI is only starting to be
integrated into language education. Third, current discourse around GenAI
centers on its technical affordances and limitations since this technology is
relatively new to language teachers and learners (Huang et al., 2023). Deeper
theoretical discussions are to be expected in the coming years (Pérez-Paredes
et al., 2025; Yang et al., 2025).
Yet, this isolation of the GenAI topic should not
be deemed as a shortcoming. Not aligning with existing topics can also mean a
potential innovation in the future learning paradigm. This is indeed happening
in language education as we have seen GenAI is positioned not only as a tool
but as a game-changer for teachers and learners across lesson planning,
learning activities design, assessment reform, and more. In order to better
leverage GenAI in innovative language education, effort is required to maximize
its transformative potential from theoretical, pedagogical, technical, and
methodological perspectives.
4.2 An Integrated Model of Innovation in CALL
This
review depicts the current landscape of innovative CALL based on a bibliometric
analysis of 1,098 articles published in eight CALL-related journals. Overall,
the analysis reveals nine core research topics over the past decade. Some of
these topics form tightly interconnected clusters that constitute the
foundations of this field, while others represent emerging topics with
potential to innovate and transform the CALL paradigm.
The framework in Figure 6 showcases the structure
and core mediating factors of innovative CALL. Existing research shows a
three-layer hierarchical model of innovation:
At the heart of this model are two mediating
constructs: Teacher Agency and Learner Agency. Teacher agency, as the most
central construct from the analysis, drives the design, adaptation, and
guidance of technological integration. Learner agency, on the other hand,
embodies engagement, self-regulation, interest, and ownership of the learning
process, serving as the internal engine through which innovative technological
experiences are processed and converted into meaningful learning gains.
Collectively, the three core layers and two
mediating factors are oriented toward the ultimate goal of fostering language
competence and broader global competencies. The value of innovation in CALL
does not reside in the technology itself but is achieved by the dynamic
interplay between agentic teachers and learners within and across each layer.
This integrative model provides a coherent conceptual map for situating diverse
research strands and for guiding future inquiry into the multifaceted nature of
innovation in CALL.
Figure
6. Multidimensional innovation ecosystem
in CALL
5. Implications and Limitations
Based
on the findings of this review, a few important implications emerge for
practice and research. First, basic educational topics such as teacher
education and course design remain stable in the CALL field; however, there is
a clear need for teachers and educators to innovate both the language learning
experiences and outcomes with emerging technologies. For example, VR offers
innovative ways to immerse learners in authentic and interactive situations,
but little is known due to its technical and resource constraints (Lee et al.,
2024). Second, Figure 5 illuminates that topics of innovation in CALL research
appear to evolve through different stages from emergence, rapid development, to
maturity, and stabilization. This cycle reminds us that scholarly attention
should be tailored to a topic’s developmental phase. For example, for
high-growth topics such as T2 and T6, more empirical research is in need to
establish foundational evidence and pedagogical frameworks. In contrast, for
well-established topics such as T7, systematic reviews and meta-analyses may
provide greater value by synthesizing existing knowledge and suggesting future
trends to further innovate the line of inquiry. Third, the eight research
topics generated from BERTopic modeling outline a roadmap for future inquiry.
These topics hold potential for synergistic integration. For instance, future
studies should explore how VR (T2) can enhance affections in L2 speaking (T0),
particularly in intercultural communicative scenarios (T3). Such cross-topic
explorations can be of great value to more forward innovative language
education.
Akin to other review studies, this review has its
own limitations. First, this review was limited to eight leading CALL-related
journals. Yet, given its interdisciplinary nature, a large quantity of research
on innovation and technology is now published in non-CALL journals such as RELC Journal. Second, as the search
strategy relied on the term “innovation,” the review may have excluded studies
that embody innovative practices or technologies, but without using this
specific term. Despite these limitations, this review offers up-to-date,
data-driven insights into the current landscape of innovation within CALL
publications. This paper provides a valuable reference for future inquiry into
technology-enhanced innovation in language education.
References
Amiryousefi,
M. (2019). The incorporation of flipped learning into conventional classes to
enhance EFL learners’ L2 speaking, L2 listening, and engagement. Innovation
in Language Learning and Teaching, 13(2), 147-161. https://doi.org/10.1080/17501229.2017.1394307
Chen,
X., Zou, D., Xie, H., Cheng, G., & Liu, C. (2022). Two decades of
artificial intelligence in education. Educational Technology &
Society, 25(1), 28-47. https://www.jstor.org/stable/48647028
Choi,
L. J., Kim, R., & Chung, S. J. (2025). Integrating
chatbot technology in language teacher education: A TPACK-based analysis of
pre-service teachers’ professional development. Language Learning & Technology, 29(1), 1–15. https://hdl.handle.net/10125/73621
Chong, S. W., & Reinders, H. (Eds.).
(2024). Innovation in language learning and teaching: The case of
England, Northern Ireland, Scotland, and Wales.
Springer. https://link.springer.com/chapter/10.1007/978-3-031-66241-6_1
Chu, Q. P., Chowdhury, R., & Le, T. T. T.
(2025). Using the UTAUT2 model to determine the factors affecting students’
acceptance of blended learning for English writing. CALL-EJ, 26(3), 24-42. https://doi.org/10.54855/callej.252632
Colpaert,
J. (2018). Transdisciplinarity revisited. Computer Assisted Language
Learning, 31(5-6), 483-489. https://doi.org/10.1080/09588221.2018.1437111
Dashtestani,
R. (2022). The winding path towards implementing digital game-based learning
(DGBL) in an educational context: The voices of pre-service teachers. Computer-Assisted
Language Learning Electronic Journal, 23(3), 70-93. https://callej.org/index.php/journal/article/view/413
Devlin,
J., Chang, M. W., Lee, K., & Toutanova, K. (2019). Bert: Pre-training of
deep bidirectional transformers for language understanding. In Proceedings of the 2019 conference of
the North American chapter of the association for computational linguistics:
human language technologies, 4171-4186. https://aclanthology.org/N19-1423/
Dizon,
G., & Gayed, J. M. (2021). Examining the impact of Grammarly on the quality
of mobile L2 writing. The JALT CALL Journal, 17(2),
74-92. https://doi.org/10.29140/jaltcall.v17n2.336
Fathi, J., Ahmadnejad, M., & Nosrati, S.
(2024). Investigating the impact of Duolingo on second language learning
achievement and self-regulated learning among French as a foreign language
learners. Computer-Assisted Language Learning Electronic Journal, 25(4),
379-413. https://callej.org/index.php/journal/article/view/477
Foroutan
Far, F., & Taghizadeh, M. (2024). Comparing the effects of digital and
non-digital gamification on EFL learners’ collocation knowledge, perceptions,
and sense of flow. Computer Assisted Language Learning, 37(7),
2083-2115. https://doi.org/10.1080/09588221.2022.2146724
Gardazi,
N. M., Daud, A., Malik, M. K., Bukhari, A., Alsahfi, T., & Alshemaimri, B.
(2025). BERT applications in natural language processing: a review. Artificial
Intelligence Review, 58(6), 1-49. https://link.springer.com/article/10.1007/s10462-025-11162-5
Hew, K. F., & Cheung, W. S. (2014). Students’
and instructors’ use of massive open online courses (MOOCs): Motivations and
challenges. Educational Research Review, 12, 45-58. https://doi.org/10.1016/j.edurev.2014.05.001
Huang,
X., Zou, D., Cheng, G., Chen, X., & Xie, H. (2023). Trends, research issues
and applications of artificial intelligence in language education. Educational
Technology & Society, 26(1), 112-131. https://www.jstor.org/stable/48707971
Hwang, Y., & Lee, S. M. (2024). ‘Can we go to
Tower Bridge?’: teaching British culture from text to immersive
realities. Innovation in Language Learning and Teaching, 1-12. https://doi.org/10.1080/17501229.2024.2439406
Kohnke,
L., Moorhouse, B. L., & Zou, D. (2023). ChatGPT for language teaching and
learning. RELC Journal, 54(2), 537-550. https://doi.org/10.1177/00336882231162868
Lai, S. L. (2023). Connecting in-class and
out-of-class learning: Integrating online resources and blog writing for EFL
proficiency enhancement. The JALT CALL Journal, 19(3),
417-440. https://doi.org/10.29140/jaltcall.v19n3.1202
Lee,
S. M. (2019). Her Story or their own stories? Digital game-based learning,
student creativity, and creative writing. ReCALL, 31(3),
238-254. https://doi.org/10.1017/S0958344019000028
Lee,
S. M., & Wu, J. G. (2023). Teaching with immersive Virtual Reality:
Perceptions of Korean trainee teachers. International Journal of
Computer-Assisted Language Learning and Teaching, 13(1), 1-14. https://www.igi-global.com/article/teaching-with-immersive-virtual-reality/334362
Lee,
S. M., & Wu, J. G. (2024). Preparing teachers for the future: Microteaching
in the immersive VR environment. ReCALL, 36(3), 251-269. https://doi.org/10.1017/S0958344024000089
Lee, S. M., Yang, Z., & Wu, J. G. (2024).
Live, play, and learn: Language learner engagement in the immersive VR
environment. Education and Information Technologies, 29(9),
10529-10550. https://doi.org/10.1007/s10639-023-12215-4
Li,
K., Peterson, M., & Wan, J. (2022). Digital gaming in the language
classroom: Student language performance, engagement, and perception. International
Journal of Computer-Assisted Language Learning and Teaching, 12(1),
1-25. https://www.igi-global.com/gateway/article/291110
Li,
M., & Zhu, W. (2017). Explaining dynamic interactions in wiki-based
collaborative writing. Language Learning
& Technology, 21(2), 96–120. https://www.lltjournal.org/item/2998
Lin,
P. (2022). Developing an intelligent tool for computer-assisted formulaic
language learning from YouTube videos. ReCALL, 34(2),
185-200. https://doi.org/10.1017/S0958344021000252
Link,
S., Mehrzad, M., & Rahimi, M. (2022). Impact of automated writing
evaluation on teacher feedback, student revision, and writing improvement. Computer
Assisted Language Learning, 35(4), 605-634. https://doi.org/10.1080/09588221.2020.1743323
Meihami, H., & Esfandiari, R. (2021).
Exploring EFL teachers’ professional identity development in a CALL teacher
preparation program. The JALT CALL Journal, 17(2),
135-157. https://doi.org/10.29140/jaltcall.v17n2.404
Pérez-Paredes,
P., Curry, N., & Ordoñana-Guillamón, C. (2025). Critical AI literacy for
applied linguistics and language education students. Journal of China
Computer-Assisted Language Learning. https://doi.org/10.1515/jccall-2025-0005
Phung,
L., Reinders, H., & Pham, V. P. H. (Eds.). (2024). Innovation in
language learning and teaching: The case of Vietnam and Cambodia. Palgrave
Macmillan. https://doi.org/10.1007/978-3-031-46080-7
Priego,
S., & Liaw, M. L. (2017). Understanding different levels of group
functionality: Activity systems analysis of an intercultural telecollaborative
multilingual digital storytelling project. Computer Assisted Language
Learning, 30(5), 368-389. http://dx.doi.org/10.1080/09588221.2017.1306567
Raman,
R., Pattnaik, D., Hughes, L., & Nedungadi, P. (2024). Unveiling the
dynamics of AI applications: A review of reviews using scientometrics and
BERTopic modeling. Journal of Innovation & Knowledge, 9(3),
100517. https://doi.org/10.1016/j.jik.2024.100517
Reinders,
H., Park, J. K., & Lee, J. S. (Eds.). (2025). Innovation in
language learning and teaching: The case of Korea. Palgrave Macmillan. https://doi.org/10.1007/978-3-031-83561-2
Song,
Y., Wen, Y., Yang, Y., & Cao, J. (2023). Developing a ‘Virtual Go mode’on a
mobile app to enhance primary students’ vocabulary learning engagement: an
exploratory study. Innovation in Language Learning and Teaching, 17(2),
354-363. https://doi.org/10.1080/17501229.2022.2047693
Teng,
M. F. (2025). Validation of vocabulary learning strategies in watching
audiovisual materials. The JALT CALL Journal, 21(2),
102936-102936. https://doi.org/10.29140/jaltcall.v21n2.102936
Tour,
E., Gindidis, M., & Newton, A. (2021). Learning digital
literacies through experiential digital storytelling in an EAL context: an
exploratory study. Innovation in Language Learning and Teaching, 15(1),
26-41. https://doi.org/10.1080/17501229.2019.1659278
Wright,
C., & Furneaux, C. (2021). ‘I Am Proud of Myself': Student Satisfaction and
Achievement on an Academic English Writing MOOC. International Journal
of Computer-Assisted Language Learning and Teaching, 11(1),
21-37. https://www.igi-global.com/gateway/journal/41023
Wu,
J. G., & Lee, S. M. (2025). Immersive literacy. In M. Pegrum., & A. Alm
(Eds.), The Palgrave Encyclopedia of Computer-Assisted Language
Learning. Palgrave. https://doi.org/10.1007/978-3-031-51447-0_264-1
Wu, J. G., & Miller, L. (2021). Raising native
cultural awareness through WeChat: a case study with Chinese EFL
students. Computer Assisted Language Learning, 34(4),
552-582. https://doi.org/10.1080/09588221.2019.1629962
Wu,
J. G., & Miller, L. (2025). Smart or sweat: The bittersweet journey of
teachers’ AI literacy. International Journal of TESOL Studies, 1-10. https://doi.org/10.58304/ijts.250124
Wu, J. G., Miller, L., Huang, Q., & Wang, M.
(2023). Learning with immersive virtual reality: An exploratory study of
Chinese college nursing students. RELC Journal, 54(3), 697-713. https://doi.org/10.1177/00336882211044860
Wu,
J. G., Teng, F., & Miller, L. (2022). Challenges in a mobile
telecollaborative project: towards a conceptual model. Australasian
Journal of Educational Technology, 38(1), 1–19. https://doi.org/10.14742/ajet.6371
Wu, J. G., Zhang, D., & Lee, S. M. (2024).
Into the brave new Metaverse: Envisaging future language teaching and
learning. IEEE Transactions on Learning Technologies, 17,
44-53. https://ieeexplore.ieee.org/document/10079929
Wu,
J. G., Zhang, D., Lee, S. M., & Xian, J. (2025). Game
on! Digital gaming and augmented reality/virtual reality in language learning.
International Journal of Computer-Assisted Language Learning and Teaching,
15(1), 1-15. https://doi.org/10.4018/IJCALLT.367601
Wu,
S. (2021). Unpacking themes of integrating telecollaboration in language
teacher education: a systematic review of 36 studies from 2009 to 2019. Computer
Assisted Language Learning, 36(7), 1265–1287. https://doi.org/10.1080/09588221.2021.1976800
Xian,
J., Wu, J. G., & Lee, S. M. (2025). Modeling the evolution of virtual
reality in nursing education: a BERTopic-based analysis of research trends and
future directions. BMC Nursing, 24(1), 1332. https://doi.org/10.1186/s12912-025-03938-5
Yang,
Z., Wu, J. G., & Xie, H. (2025). Taming Frankenstein’s monster: Ethical
considerations relating to generative artificial intelligence in
education. Asia Pacific Journal of Education, 45(4), 1330-1343. https://doi.org/10.1080/02188791.2023.2300137
You, H.-J., Yeh, H.-C., & Qi, G. Y. (2025).
Fostering intercultural sensitivity in language learning: Quality Talk in
telecollaboration. Language Learning
& Technology, 29(1), 1–26. https://doi.org/10.64152/10125/73624
Zeevy-Solovey,
O. (2024). Comparing peer, ChatGPT, and teacher corrective feedback in EFL
writing: Students' perceptions and preferences. Technology in Language
Teaching & Learning, 6(3), 1482-1482. https://doi.org/10.29140/tltl.v6n3.1482
Zhang, D., Wu, J. G., & Fu, Z. (2024). From
shy to fly: Facilitating EFL learners’ willingness to communicate with an AI
chatbot and an intelligent tutoring system. System, 127, 103501. https://doi.org/10.1016/j.system.2024.103501
Zhang, R., & Zou, D. (2022). Types, purposes,
and effectiveness of state-of-the-art technologies for second and foreign
language learning. Computer Assisted Language Learning, 35(4),
696-742. https://doi.org/10.1080/09588221.2020.1744666
Zhou,
S., & Du Preez, G. (2025). Bringing Cinderella to
spotlight? Developing grammar through ChatGPT in writing. Language Learning & Technology,
29(1), 1–25. https://doi.org/10.64152/10125/73639
Author
bio
Junjie Gavin Wu is an Assistant Professor (Tenured) and Ph.D.
Supervisor at Macao Polytechnic University, as well as the Vice President of
PacCALL and the Vice President of GloDAL. He is Editor-in-Chief of SN Social Sciences (Scopus, Springer),
Editor-in-Chief of Applied Language
Sciences, an Executive Associate Editor of Computers & Education: X Reality (ESCI, Elsevier), and an
Executive Associate Editor of Artificial
Intelligence in Language Education (Castledown). Gavin has authored around
80 English publications, with over 35 papers appearing in SSCI/SCI journals. He
has published books with Cambridge University Press, Routledge, and Springer,
and spearheaded over 10 special issues. Gavin serves on the editorial boards of
several leading journals, such as the British Journal of Educational
Technology, Interactive Learning Environments, Language Learning &
Technology, Computers and Education: Artificial Intelligence, IJCALLT,
Multimedia-Assisted Language Learning, and so on. He is also an
award-winning teacher in China.
E-mail: gavinjunjiewu@gmail.com
ORCID:
0000-0003-4937-4401
Sangmin-Michelle Lee is a professor in the Department of Metaverse at
Kyung Hee University in Korea. She is an Associate Editor of Language Learning & Technology, The JALT CALL Journal, and Applied Language Sciences. She has
published extensively on language learning in virtual reality, machine
translation, game-based learning, and digital creativity in Computer
Assisted Language Learning, ReCALL, Language Learning & Technology, British
Journal of Educational Technology, IEEE Transactions on Learning Technologies,
and Foreign Language Annals.
E-mail: sangminlee@khu.ac.kr
ORCID:
0000-0002-7686-3537
Di Zou is an associate professor in the Department of
English and Communication, The Hong Kong Polytechnic University. Her research
interests include AI in language education, flipped classroom, and second
language acquisition. She has published around 150 research papers in
international journals, conferences, and books. She serves as an Associate
Editor for Computers & Education
(SSCI IF=8.9) and Computers &
Education: X Reality, and previously for the Australasian Journal of Educational Technology. She is a member of
the editorial boards of Language Learning
& Technology, Educational
Technology & Society, and the International
Journal of Mobile Learning and Organisation.
E-mail: daisy.zou@polyu.edu.hk
ORCID:
0000-0001-8435-9739
Junhua Xian is a PhD candidate at the Faculty of Applied
Sciences, Macao Polytechnic University. His research interests include virtual
reality, STEM education, and special education and published in Educational
Technology & Society, BMC Nursing, and International Journal of
Computer-Assisted Language Learning and Teaching.
E-mail: junhua.ax@gmail.com
ORCID:
0009-0001-0585-7535