Based on correlational and factorial analysis of data collected from 384 middle and high school students in South Korea, Al–Hoorie et al. (2024) claimed the existence of a discriminant validity crisis within the L2 motivational self-system research tradition and advocated for abandoning research in this area. In this response, we critically examined the evidence presented, re-analyzed their data, and argued that their findings actually support the discriminant validity of the target scales. We also discussed issues related to the design and implementation of their study and refuted their assertion regarding a discriminant validity crisis in this field. Finally, we emphasized the necessity of prioritizing definitional validity in the ongoing methodological reforms of L2 motivation research.

Growing evidence suggests that ratings of second language (L2) speech may be influenced by perceptions of speakers’ affective states, yet the size and direction of these effects remain underexplored. To investigate these effects, 83 raters evaluated 30 speech samples using 7-point scales of four language features and ten affective states. The speech samples were 2-min videorecordings from a high-stakes speaking test. An exploratory factor analysis reduced the affect scores to three factors: assuredness, involvement, and positivity. Regression models indicated that affect variables predicted spoken language feature ratings, explaining 18–27% of the variance in scores. Assuredness and involvement corresponded with all language features, while positivity only predicted comprehensibility scores. These findings suggest that listeners’ perceptions of speakers’ affective states intertwine with their spoken language ratings to form a visual component of second-language communication. The study has implications for models of L2 speech, language pedagogy, and assessment practice.

This research report presents the development and validation of Auto Error Analyzer, a prototype web application designed to automate the calculation of accuracy and its related metrics for measuring second language (L2) production. Building on recent advancements in natural language processing (NLP) and artificial intelligence (AI), Auto Error Analyzer introduces an automated accuracy measurement component, bridging a gap in existing assessment tools, which traditionally require human judgment for accuracy evaluation. By utilizing a state-of-the-art generative AI model (Llama 3.3) for error detection, Auto Error Analyzer analyzes L2 texts efficiently and cost-effectively, producing accuracy metrics (e.g., errors per 100 words). Validation results demonstrate high agreement between the tool’s error counts and human rater judgments (r = .94), with microaverage precision and recall in error detection being high as well (.96 and .94 respectively, F1 = .95), and its T-unit and clause counts matched outputs from established tools like L2SCA. Developed under open science principles to ensure transparency and replicability, the tool aims to support researchers and educators while emphasizing the complementary role of human expertise in language assessment. The possibilities of Auto Error Analyzer for efficient and scalable error analysis, as well as its limitations in detecting context-dependent and first-language (L1)-influenced errors, are also discussed.

The love factor in the field of second language acquisition has gained considerable traction since the turn of the century. This article is the first to take a variationist perspective to investigate how multilingual coupledom affects sociolinguistic development in the second language (L2). Participants were 76 users of L2 German living in Austria, all of whom were in a romantic relationship with an Austrian partner. We analyzed the effects of multilingual coupledom on self-reported changes in learners’ use of, attitudes toward, and proficiency in standard German, the Austrian dialect variety, and first language(s), and whether (psycho-)social variables moderate this relationship. Individual differences in psychological and social variables (e.g., adaptability, Open-mindedness, length of residence, orientation toward the Austrian dialect) predicted reported changes in the sociolinguistic repertoire. Qualitative analysis revealed a blended operation of socioaffective and exposure-related factors, which helped explain why, how, and for whom multilingual coupledom affects (socio-)linguistic development.

There is evidence that learning a second language (L2) can shift cognition toward that predicted for the L2 and that this effect might vary with L2 proficiency, age of acquisition, length of immersion, etc. Here we explore the previously neglected variable of language instructional conditions. Participants categorized motion events in a triads-matching task after being trained on two novel linguistic labels highlighting (in)transitivity through one of three instructional conditions. Participants who learned the relevant knowledge under a meaning-focused instructional condition (memorizing meanings of exemplar sentences) showed a higher likelihood of categorizing based on motion (in)transitivity immediately after training than a control group; those who learned under a required rule search instructional condition showed this effect only after additional practice; while those who learned through another type of form-focused instructional condition (direct metalinguistic explanation) did not show this effect even after such practice. These differences were obtained despite the fact that the three groups were matched on awareness of the target system at the level of understanding and near-perfect performance on a grammaticality judgment task. The findings are discussed in terms of the depth of processing in instructed SLA and models of language–cognition interactions.

Fluency is an essential aspect of second language (L2) oral proficiency. Recent studies have demonstrated that L1 individual speaking style is connected to L2 fluency, suggesting that L2 speech fluency does not solely represent L2-specific skills. Furthermore, task mode (monologue vs. dialogue) has been shown to influence fluency. The present study examines the extent to which these two factors (L1 speaking style and task mode) can predict L2 speech fluency, and how such connections are modified by the learners’ L2 proficiency level. The data consist of monologic and dialogic speech samples from 50 advanced students of English in their L1 (Finnish) and L2 (English). The samples were analyzed for speed, breakdown, repair, and composite fluency. The results of multiple linear regressions demonstrated high predictive power for speed, breakdown, and composite fluency dimensions, while the model for repair fluency showed weak predictive power. The results have implications for L2 fluency research.

People are frequently concerned about the impressions they make on others (referred to as metaperceptions), but their insights are often inaccurate. Illustrating the phenomenon called the liking gap, speakers interacting in their first language (L1) and second language (L2) tend to underestimate how much they are liked by their interlocutor, and these judgments often predict their desire to engage in future interaction and collaboration. To understand the scope of this bias and its consequences, we focused on L1–L2 dyadic interaction, examining metaperception as a potential barrier to conversations between university students. We recruited 58 previously unacquainted university students to perform a 10-min academic discussion task between one L1 and one L2 speaker. Afterward, the speakers (a) assessed each other’s interpersonal liking, speaking skill, and interactional behavior; (b) provided their metaperceptions of their interlocutor’s assessments of the same dimensions; and (c) estimated their interest in future interaction with the same interlocutor. All speakers showed a reliable metaperception bias to underestimate their interpersonal liking, speaking skill, and interactional behavior. However, only L1 speakers’ desire to engage in future interaction was associated with their metaperceptions of interpersonal liking. We discuss implications of this finding for understanding and promoting academic communication.

Traditional regression models typically estimate parameters for a factor F by designating one level as a reference (intercept) and calculating slopes for other levels of F. While this approach often aligns with our research question(s), it limits direct comparisons between all pairs of levels within F and requires additional procedures for generating these comparisons. Moreover, Frequentist methods often rely on corrections (e.g., Bonferroni or Tukey), which can reduce statistical power and inflate uncertainty by mechanically widening confidence intervals. This paper demonstrates how Bayesian hierarchical models provide a robust framework for parameter estimation in the context of multiple comparisons. By leveraging entire posterior distributions, these models produce estimates for all pairwise comparisons without requiring post hoc adjustments. The hierarchical structure, combined with the use of priors, naturally incorporates shrinkage, pulling extreme estimates toward the overall mean. This regularization improves the stability and reliability of estimates, particularly in the presence of sparse or noisy data, and leads to more conservative comparisons. Bayesian models also offer a flexible framework for addressing heteroscedasticity by directly modeling variance structures and incorporating them into the posterior distribution. The result is a coherent approach to exploring differences between levels of F, where parameter estimates reflect the full uncertainty of the data.

The construct of second language (L2) utterance fluency is typically operationalized through various individual temporal features. However, in natural speech, fluency (or disfluency) is often characterized by the clustering of multiple temporal features, collectively revealing the speaker’s effort in speech production or disfluency recovery. In this study, we explore the co-occurrence patterns of disfluency features in L2 speech and their associations with speakers’ L2 oral proficiency. We initially segmented all speech samples into analysis of speech (AS)-units. Within each AS-unit, six individual fluency features were manually coded, standardized, and subsequently subjected to a hierarchical-based k-means cluster analysis to examine their co-occurrence patterns. The results revealed four distinct disfluency clusters. A subsequent qualitative analysis of disfluencies in each cluster revealed distinct distributional patterns, disfluency makeup, and communicative functions. Additionally, the proportions of different disfluency clusters were significantly influenced by speakers’ proficiency level, first language background, and their interaction. These findings carry implications for L2 speaking research in general, shedding light on the intricate nature of speech fluency and presenting an alternative approach to the operationalization of this multidimensional construct.

Al-Hoorie et al. (2024: Studies in Second Language Acquisition, 1–23) illuminate a validation crisis within the second language (L2) Motivational Self System (L2MSS), revealing empirical flaws in its current measurement. Their analysis indicates a persistent lack of discriminant validity among the system’s constructs, issuing a fundamental challenge in distinguishing the concepts. These findings, echoing previous concerns, underscore a pressing need for theoretical refinement and methodological rigor within the field, leading the authors to advocate for a temporary halt in L2 self-studies to address these issues comprehensively. This commentary discusses the call for a substantive moratorium presented by Al-Hoorie et al. (2024: Studies in Second Language Acquisition, 1–23) as a necessary step toward resolving persistent challenges in the field. By highlighting historical issues and suggesting pathways for theoretical diversification and methodological advancement, I aim to foster a productive dialogue on motivational psychology in language learning while ensuring empirical robustness.

This paper responds to Al-Hoorie, Hiver, and In’nami’s (2024) critique of the second language (L2) Motivational Self System (L2MSS) by advocating for an immediate cessation of its use in the absence of substantial revision and validation. We revisit foundational studies in the tradition, exposing critical methodological flaws that we feel undermine empirical support for the model. Further, we examine systemic factors that contributed to the largely uncritical acceptance of the model. Drawing on our own experiences, we reflect on how these dynamics have obstructed the adoption of more robust motivational theories available in psychology and education. We further caution that without a stronger emphasis on validating measurement instruments, similar distractions may continue to hinder progress in the field.

This study focused on the relative contributions of cognitive aptitudes to the incidental learning and automatization of mini-language grammar. Over three sessions, participants (N = 45; first language [L1] Polish; age range: 19–35) completed computerized training in MiniItaliano as well as tasks tapping into working memory, general intelligence, and language analytic ability (LAA). The overt aim of the training was to comprehend messages placed in the context of an Italian supermarket; however, its actual goal was to acquire the target syntactic rules (i.e., subject-verb agreement, gender agreement, word order). The final session included a surprise post-test with untimed and timed grammaticality judgment tasks (GJTs) as measures of explicit and automatized knowledge, respectively. Results showed that performance on both tasks was predicted by LAA over and above other cognitive and control measures. Additionally, scores on the untimed GJTs were related to reliance on (self-discovered) rules, while timed GJTs involved general IQ.

The current study examined the comprehension and production of classifiers, case marking, and morphological passive structures among 414 child Japanese heritage speakers (mean age = 10.01 years; range = 4.02 – 18.18). Focusing on individual differences, we extracted latent experiential factors via the Q-BEx questionnaire (De Cat, Kašćelan, Prévost, Serratrice, Tuller, Unsworth, & The Q.-Be Consortium, 2022), which were then used to predict knowledge and use of these grammatical structures. The findings reveal that: (i) experiential factors such as heritage language (HL) engagement at home and within the community modulate grammatical performance differentially from childhood through adolescence, and (ii) HL proficiency, immersion experiences, and literacy systematically predict HL grammatical outcomes. These results indicate that particular language background factors hold differential significance at distinct developmental stages and that higher proficiency, richer immersion experiences, and literacy engagement in the HL are crucial for the development of core grammatical structures.

Linguistic alignment that occurs during interaction has been found to be a useful language learning mechanism. Recent second language (L2) research on alignment has primarily focused on syntactic alignment in face-to-face oral interactions (Kim & Michel, 2023). This study expands the scope of L2 alignment research by examining pragmatic alignment in group mobile text-chat tasks conducted in Korean. Furthermore, it investigates how the source of alignment (i.e., prime, recast) and learner factors (i.e., L2 proficiency, prior knowledge of the target feature, mobile literacy, and Korean typing skills) influence the extent of L2 alignment and subsequent alignment-driven language learning. Over a period of six days, 87 Korean language learners were randomly distributed across either a prime, recast prime, or control condition and completed the following: a background survey, a pretest, four alignment tasks, two posttests, and a Korean proficiency test. During the alignment text-chat tasks, learners used the KakaoTalk mobile text-chat application to interact with two native Korean speakers who elicited Korean honorific request-making expressions in specified scenarios. The prime group received model examples prior to their production, while the recast prime group received recasts in response to non-target-like production. The learners’ use of honorific expressions was evaluated for both suppliance and accuracy across the pretests, alignment task performance, and posttests. The results revealed evidence of pragmatic alignment, with the recast prime condition demonstrating greater effects on honorific request head acts compared to the prime condition. Additionally, prior knowledge of request-making strategies facilitated L2 alignment. The implications of pragmalinguistic development through alignment-driven text-chat tasks are further discussed.

In the past decade, researchers have been increasingly interested in understanding the process of language learning, in addition to the effect of instructional interventions on L2 performance gains (i.e., learning products). One goal of such investigations is to reveal the interplay between learning conditions, processes, and outcomes where, for example, certain conditions can promote attention to the learning targets, which in turn facilitates learning. However, the statistical modeling approach taken often does not align with the conceptualization of the complex relationships between these variables. Thus, in this paper, we introduce mediation analysis to SLA research. We offer a step-by-step, contextualized tutorial on the practical application of mediation analysis in three different research scenarios, each addressing a different research design using either simulated or open-source datasets. Our overall goal is to promote the use of statistical techniques that are consistent with the theorization of language learning processes as mediators.

Al-Hoorie, Hiver, and In’nami (2024) offer compelling arguments for why L2 motivational self-system research is currently in a state of validation crisis. Seeking a constructive resolution to the crisis, in this response we argue that two fundamental conditions are needed for the field to emerge stronger: psychological readiness and methodological maturity. For psychological readiness, we call for a reframing of the “crisis” narrative. We highlight the need to value controversy, to normalize failure and (self-)correction, and to resist the allure of novelty. For methodological maturity, we suggest that an argument-based approach to validation can provide a constructive solution to current controversies. We present an integrated framework that can guide systematic validation efforts, and we demonstrate its application using a recent validation study as an example.

This study investigates how second language (L2) learners engage in prediction based on their processing goals. While prediction is a prominent feature of human sentence comprehension in first–language speakers, it remains less understood when and how L2 learners engage in predictive processing. By conducting a visual–world eye–tracking experiment involving Chinese–speaking L2 learners of Korean, we tested the hypothesis that L2 learners determine whether to engage in prediction by evaluating the costs and benefits of anticipatory processing. The experiment specifically focused on the impact of a top–down comprehension goal for L2 learners’ predictive use of an honorific form in Korean by providing them with different types of task instruction. Our results indicated that all groups engaged in predictive processing in early and entire predictive regions. However, in the late predictive region, L2 learners presented with a prediction–oriented task, but not those with a simple comprehension task, actively generated expectations about the honorific status of an upcoming referent. These findings lend support to the utility account of L2 prediction, suggesting that L2 learners’ engagement in prediction depends on their current goals and strategies for processing efficiency.

Word processing during reading is known to be influenced by lexical features, especially word length, frequency, and predictability. This study examined the relative importance of these features in word processing during second language (L2) English reading. We used data from an eye-tracking corpus and applied a machine-learning approach to model word-level eye-tracking measures and identify key predictors. Predictors comprised several lexical features, including length, frequency, and predictability (e.g., surprisal). Additionally, sentence, passage, and reader characteristics were considered for comparison. The analysis found that word length was the most important variable across several eye-tracking measures. However, for certain measures, word frequency and predictability were more important than length, and in some cases, reader characteristics such as proficiency were more significant than lexical features. These findings highlight the complexity of word processing during reading, the shared processes between first language (L1) and L2 reading, and their potential to refine models of eye-movement control.

Various theories have been proposed in the field of second language (L2) sentence processing research and have significantly advanced our understanding of the mechanisms underlying L2 sentence interpretation processes. However, many existing theories have only been formulated verbally, and little progress has been made towards formal modelling. Formal modelling offers several advantages, including enhancing the clarity and verifiability of theoretical claims. This paper aims to address this gap in the literature by introducing formal computational modelling and demonstrating its application in L2 sentence processing research. Through practical demonstrations, the paper also emphasises the importance of formal modelling in the formulation and development of theory.

Skill acquisition theory conceptualizes second language (L2) learning in three stages (declarative, procedural, and automatic), yet competing theoretical models with fewer stages also exist, and the number of stages has never actually been tested. We tested the validity of the three-stage model by investigating the number and nature of learning stages in L2 skill acquisition. Seventy-three participants deliberately learned grammar and vocabulary of a miniature language through explicit-deductive instruction. They systematically practiced comprehending the language until their accuracy and speed of performance did not improve anymore. Participants received a battery of tests assessing individual differences in their declarative and procedural learning abilities. We first applied hidden Markov modeling to participants’ reaction time data (obtained from the language practice) to compare rival hypotheses on the number of stages in L2 skill acquisition. We then examined which cognitive variables predicted participants’ performances (accuracy and speed) in each stage. Our results indicated that participants indeed acquired L2 skills in three stages and that their performance correlated initially with declarative learning ability, but there was a tendency for procedural learning ability to take over in the later stages. Our findings provide the first formal evidence for the influential three-stage model of L2 skill acquisition.