Language, deeply intertwined with our cultural, social, and psychological facets, changes dynamically over time through the constant interaction of these elements. Among these linguistic phenomena, second language learning is a complex process undertaken by learners from diverse backgrounds, each with different native languages, experiences, personalities, motivations, and learning objectives. Complex Dynamic Systems Theory (CDST) provides a theoretical framework for reconsidering and re-examining second language learning from a transdisciplinary perspective, transcending the boundaries of academic fields. This paper draws on CDST to explore the nature of second language learning, the processes through which it unfolds, and the reasons why some individuals progress while others in the same educational environment do not. This paper begins by defining a “complex system” and explaining its characteristics and associated phenomena. Subsequently, I will explain a research project my co-research and I have conducted over the past 15 years. This project investigates the development of university students' second language writing skills through the repetition of a writing task. To date, we have collected over 18,000 English writing samples from more than 600 university students. The findings from this project can be consolidated into two major aspects: the long-term development of L2 writing, and changes in learner psychology, particularly concerning learner agency. Regarding the former, we investigated how task repetition affects the development of L2 writing skills, identifying the conditions for phase shifts and long-term developmental patterns. In relation to the latter, we examined how changes in learner agency interact with the L2 writing system, yielding various insights into the influence of an agentive attitude towards English learning on the development of writing skills. Based on the lessons learned from this research project, this paper will conclude by considering second language learning from a new perspective.

This paper explored shadowing as an oral practice to improve second language (L2) speaking skills, drawing on cognitive brain systems such as the working memory and Levelt's (1989) production model. The 2023 National Achievement Tests revealed that Japanese junior high school students scored low on English speaking tests, implying that developing L2 fluency in English classrooms is quite challenging in Japan. Speaking in an L2 is difficult due to its feature of simultaneous multiple processing and the lack of automatization in speech processes. Shadowing, involving listening and the immediate repetition of incoming heard speech, is proposed as an effective practice to enhance L2 speaking abilities (Kadota, 2012, 2015, 2018, 2019). The paper discussed how shadowing utilizes and strengthens the phonological loop in working memory and how it enhances the functions of the articulator and the formulator in speech production. Previous studies have shown that shadowing can improve phoneme perception, pronunciation, and faster articulation by automating speech perception and some production processes. Finally, this paper introduced a seven-step shadowing training method, which includes listening, mumbling, parallel reading, prosody shadowing, content shadowing, role-playing reading aloud, and retelling. This structured approach aims to develop speaking skills through holistic chunk learning, analytic rule learning, and automatic manipulation.

Language acquisition or learning is inherently social, relying on the effective functioning of the social brain system. This article first focuses on three essential mechanisms of the social brain that facilitate second language learning: 1) Mentalizing, 2) Mirroring, and 3) Recognizing a Gaze and Constructing Joint Attention:

1) Mentalizing: This refers to interpreting others' mental states, such as viewpoints and intentions. This ability enhances communication by enabling learners to grasp intentions or implied meanings in conversation, which is crucial for mastering L2 communication. Tasks to develop mentalizing in language education (e.g., task-based learning) which challenge students to consider their peers' perspectives and viewpoints.

2) Human Mirroring System: The mirroring system supports the imitation of visual gestures as well as speech sounds, both fundamental for communicative fluency. Engaging this system through methods such as self-contained and interactive oral shadowing may help learners acquire segmental and prosodic speech sounds, potentially leading to neural coupling or entrainment between students.

3) Recognizing Gaze and Constructing Joint Attention: This mechanism is pivotal for learning through social interaction by activating the inferior frontal gyrus in the right hemisphere. It enables learners to follow another's gaze and jointly focus on third-party objects and events, linking verbal to non-verbal information. This skill is particularly important for L1 and L2 vocabulary acquisition. Activities that promote joint attention result in collaborative tasks requiring shared focus on external objects.

In this article, I propose the concept of Social Brain Interactional Competence (SBIC) in L2 acquisition and present an overall picture of how the learners' social brain systems are involved in its attainment. However, in order to acquire SBIC, it is essential to practice I.P.O.M., which represents the key points of L2 acquisition, and achieving the minimum level of cognitive automaticity of language control appears to be a prerequisite for effective social brain interaction.

The idea that emotion is inseparably fundamental to cognition has been theorized and supported in multiple academic disciplines including philosophy, psychiatry, neuroscience, and second language acquisition. Emotion, as a socially constructed phenomenon, is also integral to socio-emotional literacy, underscoring its close relationship with both cognition and social interaction. This paper overviews three evolving hypotheses on micro-level emotions in foreign language learning. The first hypothesis is the Emotion-Involved Processing Hypothesis (EIPH), according to which emotional elaboration in semantic processing results in stronger incidental memory traces (i.e., is deeper) than semantic processing without elicited emotion involvement. The EIPH is supported by multiple empirical studies. The next question is what kind of emotion should be involved for better language learning, leading to the second abductive hypothesis: the Deep Positivity Hypothesis (DPH), according to which deeper/semantic processing and learning are facilitated more by positive emotions than negative emotions. The DPH is not only empirically tested but also is harmonious with the flourishing positive psychology with valuable practical insights in second language acquisition. Are positive emotions the panacea, then? The answer is not always yes. Based on the free energy principle, the Deep Epistemic Emotion Hypothesis (DEEH) has been proposed, according to which higher order intellectual acquisition would be facilitated more by epistemic emotions such as intellectual surprise and curiosity than hedonic emotions. The DEEH is supported empirically as well as pseudo-autoethnographically. Collectively, these evolving hypotheses provide theoretical and practical insights for understanding and making the most of the dynamic interplay between emotion and cognition in language learning.

The Japan Second Language Association (J-SLA), which publishes the journal Second Language, has defined the nature of Second Language Acquisition (SLA) research in its founding principles. The 24th International Conference of the Japan Second Language Association (J-SLA2024) held a symposium where three researchers presented their proposals on what constitutes research that aligns with this definition, and discussions followed based on those proposals. This chapter presents their proposals and the subsequent discussion. Each paper critically reviews past SLA research and offers suggestions regarding the direction SLA research should take as a scientific discipline.

In the field of second language acquisition, linguistic knowledge has traditionally been classified into two dichotomous categories: what can be utilized consciously (hereafter, conscious knowledge) and unconsciously (unconscious knowledge), and the distinction has been a focus of active debates for decades. This paper broadly introduces such frameworks concerning conscious and unconscious knowledge in second language acquisition, highlighting the historical developments in this area of research. Additionally, these traditional frameworks are examined to show that the dichotomies presented so far seem to be about the same constructs but actually are different, and that different interpretations and predictions of the same phenomenon can arise depending on the framework on which they are based. In the end, I suggest that theories about the consciousness inherent in knowledge are potentially relevant to research based on any linguistic theory, as long as it studies a second language. It is then shown that this theoretical study of consciousness requires a logic of inquiry similar to that of linguistic theory, and future perspectives are discussed.

This paper provides an overview of conceptual issues related to the hypotheses proposed in generative approaches to second language acquisition (GenSLA). Generative Grammar refers to an explicit grammatical system, wherein only explicitly defined theoretical mechanisms are assumed, and numerous highly falsifiable explanatory hypotheses have been proposed. After introducing some of these efforts, this paper critically examines the hypotheses put forward in GenSLA to date. In particular, it argues that many recent hypotheses tend to have a descriptive nature and fail to pursue deeper explanations. These hypotheses also suffer from fundamental issues, such as nonexplicit claims and lack of testability. Furthermore, taking the Representational Deficit Hypothesis and the Missing Surface Inflection Hypothesis as examples, the paper discusses the importance of constructing explicit explanatory hypotheses based on an appropriate linguistic theory for the observed descriptive generalizations.

In this paper, I will review the current status of research on second language (L2) vocabulary acquisition and discuss potential directions for future research. Specifically, I will address issues related to: (1) the model-free nature of research, (2) the definition of vocabulary acquisition, and (3) the acquisition of single words vs. multi-word items. First, Meara (1997) argued that the literature on L2 vocabulary acquisition is “remarkably model-free” (p. 111), hindering our understanding of the vocabulary learning process. I will discuss what progress has been made in this area. Additionally, several studies have explored how findings from cognitive psychology (e.g., retrieval effect, distributed practice effect) apply to L2 vocabulary learning. I will examine the contributions and challenges of integrating cognitive psychology into L2 vocabulary research. Second, vocabulary knowledge encompasses at least two dimensions: vocabulary size (i.e., How many words do learners know?) and vocabulary depth (i.e., What knowledge do learners have about each word?). Most existing studies have operationalized vocabulary acquisition as initial form-meaning mapping, ignoring the depth of word knowledge. In the past two decades, however, attempts have been made to examine factors that influence the learning of vocabulary depth. Recent studies have also investigated both declarative and non-declarative vocabulary knowledge through measures such as priming and eye-tracking. Third, although early research primarily focused on the acquisition of single words, recent studies have underscored the significance of multi-word items (e.g.., idioms, collocations, phrasal verbs) in the acquisition, processing, and use of L2. This has prompted researchers to examine the acquisition of not only single words but also multi-word items. I will discuss the value and potential challenges of research on the acquisition of multi-word items.

This study investigates the second language (L2) acquisition of specificity marked by the Japanese plural -tati, within the framework of the Feature Reassembly Hypothesis (FRH) (Lardiere, 2008, 2009). Japanese -tati optionally expresses plurality, but it is also constrained by specificity—a noun suffixed with -tati is obligatorily specific hence incompatible with contexts with no specific referent. We examine whether L2 learners can acquire the specificity constraint of -tati and whether a similar property in the first language (L1) facilitates acquisition. We target L2 Japanese learners with two different L1s: Korean, which has a plural marker with a similar specific reading (-tul); and English, whose counterpart (-s) does not concern specificity. The results of an acceptability judgement task showed that native Japanese speakers are sensitive to the specificity constraint, and that the L2 learners' knowledge of the target property is, despite their descriptively target-like response pattern, not as evident, with no concrete evidence of L1 effects. Nevertheless, an individual analysis confirmed that some advanced L2 learners in each L1 group have robust knowledge of the constraint, suggesting that L2 acquisition of this constraint is at least possible. We attribute the considerable difficulty and the absence of L1-related difference in the L2 acquisition of -tati to opacity of the form-meaning mapping due to its multiple meanings of plurality, specificity, and associativity (‘x and others’ reading). That is, a single form with more than one meaning can increase the difficulty of establishing the form-meaning correlations. In terms of the FRH, a prerequisite for mapping an L1 meaning to an L2 form and for subsequent feature reassembly is that the form-meaning correlation must be detected (otherwise, neither L1 transfer nor L2 acquisition will occur). Therefore, we argue that the challenge lies with the initial feature mapping, and that the findings can be explained within the FRH.