The Second Language Research Forum (SLRF) 2025 was held at Northern Arizona University (NAU) in Flagstaff, Arizona, on September 25–28, 2025. There were 316 participants from 12 countries and 95 universities who came to Flagstaff, Arizona, for three days of inspiring presentations and thoughtful discussions. It was thrilling to reconnect with our colleagues and friends and to learn about the very latest in the field of second language research.
For those who might be unfamiliar, SLRF began in 1977 at the University of California Los Angeles (UCLA). Between 1977 and 1987, it was held seven times – five times at UCLA and twice at the University of Southern California (USC). Since 1988, the conference has been hosted by a wide variety of universities in both the United States and Canada. Historically, the hosts in any given year have been tasked with soliciting proposals from and selecting the following year’s host. This fairly informal tradition has generally worked. However, due to the pandemic, the hosts of SLRF 2020 (Vanderbilt University) understandably received no proposals to host in 2021. Consequently, for several years, the conference was left dormant.
But SLRF was not forgotten. As in-person events began to return, a number of scholars began to wonder whether, when, and where SLRF might return. Then, in spring 2024, we (the Applied Linguistics program at NAU) began preliminary discussions led by Scott Jarvis to explore the possibility of hosting SLRF 2025. After conferring with organizers at the two previous host institutions (Vanderbilt University and Michigan State University) and exploring some of the logistics involved, we decided to host SLRF 2025 and then immediately formed an organizing committee. Given SLRF’s tradition of being student-run, a team of three PhD students immediately volunteered to serve as conference Co-Chairs: Francesca Grixoni, Taehyeong (Terry) Kim, and Shujing Zhao. Three NAU faculty members, Bill Crawford, Scott Jarvis, and Luke Plonsky, agreed to serve as faculty advisors.
The program and related activities, which we describe below, were inspiring and full of the energy and insight that has characterized this event for decades. Our sincere thanks to all the volunteers, presenters, and attendees who came. We are also very grateful for the funding we received from a number of units across NAU as well as the generous support from Cambridge University Press, John Benjamins, Duolingo, and Languages.
1. Plenaries
SLRF 2025 had two excellent plenary speakers who presented on two important and enduring topics in the field of second language acquisition (SLA). The first – ISLA as a team name: Exploring its inclusive and equitable future for second language education by Dr. Masatoshi Sato (Universidad Andrés Bello) – proposed two future directions of instructed second language acquisition (ISLA) research. Dr. Sato first advocated to move beyond ISLA research that adopts a primarily cognitive and quantitative perspective to include a wider range of perspectives (i.e., social, affective, multilingualism). The second direction advocated for teachers to have a more prominent voice in researcher–practice relationships. Dr. Sato argued that these two directions can be achieved by focusing on the broad concept of second language (L2) use in a variety of classroom contexts as opposed to more traditional explorations of (grammatical and lexical) complexity, accuracy, and fluency. Dr. Sato also advocated for journals to promote work that investigates how pedagogical recommendations can be expressly investigated in classroom contexts and encouraged journals to include work that incorporates the perspectives of those who are more directly involved in classroom practices.
The second plenary – Grammatical complexity in L2 writing development: Contrasting oral vs. literate complexity across registers and developmental levels by Dr. Douglas Biber (Northern Arizona University) – presented a perspective of grammatical complexity that identifies grammatical complexity features which contrast oral and literate contexts of language use. This perspective illustrates how learners’ acquisition of L2 writing complexity moves from those features associated with oral language to those associated with literate language. Dr. Biber then showed how some current work with colleagues tracks the development of these features from English specific writing classes, through undergraduate and graduate writing and then to specific academic disciplines. This approach to the investigation of grammatical complexity is especially relevant to those interested in grammatical development because this research identifies functionally and linguistically interpretable aspects of performance as opposed to more omnibus measures such as T-unit length.
2. Papers
With 189 individual paper presentations, SLRF 2025 showcased the breadth of current L2 research. To highlight what these presentations collectively reveal about emerging directions in the field, a list of words that appeared in abstracts was compiled. Several notable thematic trends were identified across the SLRF 2025 program, with numbers in parentheses indicating how many presentations addressed each theme or context.
Presentations were situated primarily in university (41) and classroom (35) contexts, which highlighted the centrality of both higher education and instructional settings to current L2 research. Beyond research contexts, the breadth of target and native languages represented across the program illustrated the diversity of the learner population, with English (94), Spanish (36), Chinese (23), Japanese (20), Korean (14), and French (14) appearing most frequently. The program further demonstrated sustained attention to both perceptions (23) and learner performance or production (29 and 21, respectively). Research also spanned multiple L2 skill domains, most notably writing (34), reading (17), and speaking (16). Finally, across strands, recurrent themes emerged around task (64), processing (34), instruction (31), speech (30), vocabulary (30), (individual) differences (33), complexity (26), engagement (24), accuracy (26), feedback (19), and assessment (17), reflecting the diverse and interconnected directions that characterize current L2 research.
This breadth of topics was well represented by both senior scholars as well as early career researchers and graduate students in the field, often through collaborative work. HeeJoon Choi, Yingliang Liu, and YouJin Kim, for example, compared the effects of digital multimodal composing (DMC) and traditional monomodal writing on vocabulary learning and reading comprehension in integrated reading-to-write tasks among Chinese EFL learners. The study found that writing modality influenced both vocabulary acquisition and reading comprehension, suggesting potential advantages of DMC for integrated literacy development and informing pedagogical approaches to multimodal integration in L2 instruction. Yoonseo Kim and Daniel Isbell likewise exemplify the junior + senior scholar collaborations found at SLRF. Their talk presented an item-level method for scoring graph-description tasks with large language models. Comparing several state-of-the-art LLMs to human raters, they found solid alignment – especially on elementary items – and clear differences in rater severity across models. Overall, the approach shows promise for augmenting human scoring by capturing graph-specific content and delivering targeted, item-level feedback.
Methodologically, the studies drew on both qualitative (25) and quantitative (16) traditions, and the emphasis on methods (19) in the program further underscored the field’s growing commitment to methodological rigor and innovation in L2 research. Tove Larsson and Gregory R. Hancock discussed methods for testing similarity rather than difference between groups. By explaining the use of equivalence testing and mean structure modeling, they encouraged researchers to complement traditional null hypothesis testing with approaches that can formally establish practical equivalence, broadening the scope of quantitative inference in the field. Several papers showcased transparent, audit-trailed qualitative designs; for example, Carlo Cinaglia’s narrative-inquiry study, framed by the investment construct, traced how undergraduates negotiated identities and commitment across classroom and out-of-class spaces, triangulating observations, structured and semi-structured interviews, and written reflections with clear analytic steps. Seventeen studies explicitly adopted mixed-methods designs; among these, Tina Tan and Lydia Ahn presented a convergent project on second-generation immigrants’ shared language fluency and family communication, whose qualitative arm used Robinson’s Structured Tabular Thematic Analysis to synthesize large-N short-text responses into inspectable theme tables linking shared fluency to communication patterns and emotion socialization, which in turn informed measurement choices in the quantitative arm.
On par with the field’s rapid evolution, 15 presentations focused on the role and application of AI and digital tools in L2 research and pedagogy. One example is the study by Dayoung Joo and Diane Belcher, which examined how learners actually use generative AI during academic reading. Framed by sociocultural theory, their qualitative cases show ChatGPT moving beyond simple translation to serve as a scaffold for both bottom-up processing (e.g., vocabulary decoding) and top-down work (activating background knowledge), while also reshaping students’ study practices and views of assistance.
Another example of AI’s instructional utility came from Jeffrey Stewart, Christopher Nicklin, Laurence Anthony, Aaron Batty, Keita Nakamura, Stuart McLean, and Kanako Tomaru, who asked: Can we reliably score meaning-recall vocabulary tests using AI? Comparing LLM scoring with human ratings, they reported encouraging alignment and systematic rater-severity differences by model. They also previewed an AI-scored meaning-recall platform that lowers item-authoring workload (by auto-expanding acceptable answers) and speeds feedback – while keeping humans in the loop for quality control. Following a four-year hiatus, SLRF 2025 underscored the field’s renewed energy, range, and methodological strength – reaching new heights in L2 research.
3. Workshop, roundtables, and posters
SLRF 2025 aimed to create an interactive and engaging conference experience and provide attendees with diverse opportunities to exchange ideas and connect with each other. Therefore, in addition to the traditional paper sessions, we also included roundtable and poster presentations, as well as a hands-on preconference workshop, designed to foster discussion and inspiration.
The preconference workshop took place on the afternoon of September 25, the day before the main conference. We were honored to welcome Dr. Naoko Taguchi (Northern Arizona University) and Dr. Wenhao Diao (University of Arizona) as the workshop facilitators. They delivered a dynamic and practical session on Mixed-Methods Research (MMR) in second language acquisition. To create an interactive experience where the speakers could engage closely with all participants, attendance was capped at 30. The workshop included an overview of MMR, a sample study presentation, and guided activities on identifying mixed-methods designs, conducting thematic and discourse analysis, and developing participants’ own MMR projects. Attendees described the workshop as both insightful and inspiring and emphasized how valuable it was to exchange feedback with peers in such a collaborative environment, which was exactly the kind of hands-on experience we had hoped to create.
Throughout the main conference, we continued to prioritize interaction. Alongside individual paper presentations, SLRF 2025 featured roundtable sessions where presenters shared work in progress and engaged in constructive conversations with fellow researchers. Each roundtable included three to four presenters grouped by research area, facilitating a deeper exchange of expertise. Presenters had ten minutes to introduce their work, followed by five minutes of Q&A. There were also dedicated timekeepers helping facilitate transitions. We also hosted poster sessions, including nine groups of presenters whose work was showcased during scheduled breaks, so that participants could engage fully without overlapping with other sessions. The poster sessions were intentionally small in scale and encouraged closer and more genuine conversations between presenters and attendees. The result was an energetic atmosphere where researchers could connect meaningfully over shared interests and new ideas.
Overall, SLRF 2025 successfully brought together the L2 research community in a way that emphasized collaboration and engagement. This year’s event also showed that meaningful academic exchange happens through conversations that inspire the next step of inquiry.
4. Social events
Since SLRF had been dormant since 2020, we designed the 2025 social program to help reknit the community that many attendees identify as the conference’s trademark – while also showcasing the beauty of Northern Arizona. The conference dinner used mixed seating, which put graduate students, early-career researchers, and senior scholars in conversation at the same tables and led to easy mentoring and new collaborations. An informal Discord backchannel handled quick logistics and meet-ups without adding to the schedule. Lively coffee breaks and the poster session provided further space for ongoing discussions. For those remaining in Flagstaff, optional Sunday outings to the Grand Canyon and Sedona extended conversations in a different setting. A planned evening at Lowell Observatory had to be cancelled due to cloud cover. Overall, the social elements fostered community and exchange while keeping the spotlight on the research.
5. The future of SLRF
SLRF 2026 will be hosted by Université de Montréal in cooperation with the Montreal Institute of Second Language Acquisition and the three other universities in Montreal that have graduate programs with concentrations in or closely aligned with Second Language Acquisition (i.e., Concordia University, McGill University, and Université du Québec à Montréal). As always, the conference will be run by an organizing committee made up of graduate students in relevant fields of specialization. The SLRF 2026 organizing committee is supported by a group of faculty mentors led by Dr. Gabriel Michaud.
A potential host for SLRF 2027 is also currently under consideration. Following past tradition, the student organizers for SLRF 2026 in Montreal will select the host for the following year. Looking beyond 2027, and recognizing that the SLRF conference has not been held in 10 of its 49 years due to the absence of hosting proposals, we have initiated informal discussions with colleagues at other institutions to explore the creation of a North American consortium of SLA-related programs that would oversee SLRF and rotate hosting responsibilities. We believe that such a consortium would have three main benefits:
1. It would provide structure and predictability, enabling member institutions to plan several years in advance for their turn to host SLRF.
2. It would help ensure the continuity of SLRF, preventing future lapses in the conference’s history.
3. It would guarantee that each annual SLRF is hosted by an institution with a strong SLA-related program and a sufficient number of PhD students to serve as conference organizers.
These plans are still in their infancy. Regardless of how future SLRF hosts are chosen, we remain committed to putting in place the structures and mechanisms needed to ensure that this valuable conference – one that many scholars describe as their preferred academic gathering in North America – will continue to thrive for many years to come.
