How a Usage-Based Approach Promotes Conceptual Development and Natural Use of Japanese Passives: Evidence from Concept-Based Language Instruction

Abstract

L1 transfer is well-attested in SLA; negative transfer is common when learners encounter a typologically distinct language. English-speaking learners often struggle with Japanese passives, which differ significantly from English passives both conceptually and grammatically. While English passives primarily defocus the agent, Japanese passives serve multiple semantic and discourse functions, often maintaining a focus on (and empathy toward) the experiencer. This small study examines how conceptual understandings drawn from usage-based (UB) analyses influence the acquisition of Japanese passives. Using corpus studies and acquisition research as a foundation, we developed concept-based language instruction (C-BLI) integrating UB-focused concepts. Our analysis of students’ oral languaging, gesture, and story-writing data from an immediate post-test and two delayed (3 weeks and 6 months post-instruction) post-tests show individual differences and demonstrate how a UB-based C-BLI approach facilitated developmental processes in Japanese over time; students improved their grasp of concepts taught via multi-modal materials, including visual materializations of concepts and ocean wave gestures. Conceptual and linguistic development were evidenced via oral languaging and story-writing. The most frequently used passive verb was iu ‘say,’ which has been found to be often passivized in L1 speakers’ production and previous SLA research. Findings contribute to broader discussions of how conceptual restructuring may affect L2 acquisition of complex grammatical constructions.

1. Introduction

Negative transfer from a student’s first language (L1) is well attested in second language (L2) acquisition. Learners of L2 Japanese whose L1s are typologically distant, such as English and Chinese, face significant challenges acquiring many Japanese constructions. The passive construction is one such example. Whereas many language learners’ L1 passives defocus the agent, Japanese passives serve multiple semantic and discourse functions. They often foreground the experiencer’s perspective by showing an action’s impact on the experiencer, potentially evoking empathy. Building on corpus studies and acquisition research (Iori, 2012; Maeda, 2011; Minami, 2021; Mitsugi, 2018; Mitsugi & Fukada, 2022; Taguchi, 2001), we developed usage-based (UB) instructional materials within the framework of concept-based language instruction (C-BLI, Gal’perin, 1992). These materials employed visual materializations of the concepts, called SCOBAs (schema for an orienting basis of action) along with pedagogical gestures and tasks such as story-writing and oral languaging (language-related verbalization; see Swain, 2006). The goal of this approach was to foster the development of embodied UB L2 knowledge.

The present study is an extension of Masuda and Ohta (forthcoming), which examined the impact of a C-BLI unit on Japanese passives in an intact classroom, considering data from nine intermediate-level learners’ immediate and three-week-delayed post-tests. On the delayed post-test, learners’ oral languaging showed deeper understandings of passive discourse functions, increased use of ni marking the source of impact, and stronger awareness of empathy toward the experiencer. Students’ use of passives also increased in the delayed post-test story-writing, as compared to the immediate post-test, with seven of nine students maintaining or increasing their usage. These results, along with a stable TLU (Target-Like Use,1 Pica, 1983/2008) of 69% and 70%, suggest durable learning.

The present study extends Masuda and Ohta (forthcoming) by adding a 6-month-delayed post-test. The present study also includes qualitative analysis of learners’ oral languaging, which was not included in Masuda and Ohta (forthcoming). Extending this study works to address a limitation of previous C-BLI studies, which tend to have very short study duration, with the longest being one semester in length (e.g., Buescher & Strauss, 2015; Masuda & Labarca, 2015, 2018; Lantolf & Tsai, 2018; Poehner & Infante, 2019; Kissling & Muthusamy, 2022).

2. Background

2.1. What Are Japanese Passives and Why Are They Challenging for L2 Learners?

Japanese passives are known to be difficult for L2 learners to acquire. Challenges range from structural complexity (i.e., challenges in making the morphological form) to learners’ tendency to avoid passives by using active sentences, even in contexts where passives are decidedly more natural (e.g., Kawaguchi, 2005; Tanaka, 2010). More specifically, structural differences between English and Japanese passives hinder L2 learners, as do the empathy-related and discourse meanings associated with the Japanese passive. While English passives are primarily used to signal a perspective shift (Langacker, 2004),2 Japanese passives are strongly influenced by empathy and discourse factors, as discussed later in this section.3 Avoidance is particularly evident among English L1 speakers, who generally transfer English syntactic patterns into their Japanese passive usage without considering discourse factors.

Japanese passives are typically classified into two or three types, depending on theoretical orientation. Murakami (2022)4 explains that passives may be classified based on features such as animacy of the subject (animate or inanimate), the syntactic correspondence between passive and active sentences (direct, indirect, or possessive), and semantic aspects of impact (neutral or adversative). These features can be overlapping and not discrete. Since meaning-making is central to sociocultural theory, we use the terms affective passive (also called the direct passive), adversative passive (also called the indirect passive or possessive passive), and non-human subject passive. Both the affective and adversative passives encode the affective impact of an action on the experiencer. The affective passive can be used to express a positive or a negative impact, depending on the semantics of the verb and the context. Like passives in English, an affective passive sentence can be transformed into an active sentence. The adversative passive places the experiencer in subject position and uses the passive voice to encode a situation impacting them; adversative passives have no direct translation into English, as will be demonstrated below. The non-human subject passive is distinguished by a neutral tone, a requirement that its non-human subject be thematically significant, and the usage of different grammatical particles to mark the source of the action. We provide further explanation of these types of passives, from a UB perspective, in a later section of the paper where we describe our instructional materials.

Japanese affective and adversative passives are challenging for L2 Japanese learners for semantic and discourse reasons; they express the impact of a situation on the experiencer. English generally uses active sentences for this purpose, and English-dominant Japanese language learners tend to translate this thinking into their Japanese. Example (1) displays an active Japanese sentence that, while natural in its English translation, is unnatural in Japanese, because, except for the negative connotation of the overall situation, it does not otherwise encode a negative impact on the speaker, who was reporting her perspective.

(1)	Active sentence:5
	?Densha de	dareka ga	watashi no ashi o	fun-da.
	train LOC	someone NOM	I GEN foot ACC	step on-PST
	‘On the train, someone stepped on my foot.’

Japanese L1 speakers mostly use passives in this sort of situation, as demonstrated by corpus analyses of Japanese usage (Maeda, 2011; Iori, 2012). This is because Japanese favors what Kuno (1978) terms a hierarchy of empathy, whereby the speaker and in-group participants rank highest and are more likely to be selected as grammatical subjects and centers of perspective. The passive construction thus enables alignment between grammatical subjecthood and discourse perspective. In general, while English tends to foreground agents, Japanese tends to foreground the subject/speaker, whose perspective, as the experiencer, is expressed via the passive verb. These differences are grounded in different construal patterns of English and Japanese. Construal refers to how events are framed by a language. By framing utterances from the experiencer’s perspective, Japanese usage aligns with the notion of subjective construal, while English’s agent-focused orientation is based in objective construal (Ikegami, 2005, 2016; Ikegami & Moriya, 2009). Subjective construal is an experiencer-focused, subjectivity-prominent perspective. English-background learners of Japanese often under-use passives because they tend to frame events objectively, relying on direct translation of English active sentences. However, directly transforming a Japanese sentence from active to passive often results in odd usage in Japanese, as shown by Example (2), which transforms Example (1) into an affective passive sentence. This sentence is possible, but awkward, in Japanese.

(2)	Affective passive with English-transferred thinking:
	?Densha de	watashi no ashi ga	fum-are-ta.
	train LOC	I GEN foot NOM	step on-PASS-PST
	‘On the train, my foot was stepped on.’ (Okay English, unnatural Japanese.)

There are contexts in which this Japanese sentence would be natural; it could answer questions about what was stepped on or whose foot was stepped on, for example. However, for expressing, in general, what happened to the protagonist, the sentence falls short because the experiencer is not the subject of the sentence. To best describe this situation, moving the experiencer to subject position and following this with expression of the situation using the passive is most natural. This results in the adversative passive sentence shown in Example (3), which has the (unstated) experiencer in subject position:

(3)	Adversative passive, which expresses the experiencer’s perspective:
	Densha de	ashi o	fum-are-ta.
	train LOC	foot ACC	step on-PASS-PST
	‘On the train, [somebody] stepped on [my] foot.’ (The underlined verb here is in the passive form in the Japanese sentence. The words in brackets are implied in the Japanese sentence.)

English has no passive equivalent for Example (3). Here, the Japanese passive foregrounds the experiencer. The experiencer is not stated, because they are easily inferred from the context. The adversative passive most naturally conveys the empathy- and experiencer-centered subjective framing of utterances that is characteristic of Japanese.

For L2 learners, Japanese passives present a twofold challenge. First, as shown by the previous examples, passives in Japanese are often used in ways that cannot be directly translated into English. Second, the conceptual perspective that motivates the use of passive-or-not differs across languages. Learning to use language in this new way requires learners to shift from the agent-focused orientation of English, to learn a new way of seeing things. To use Japanese effectively, students need to learn to construe events subjectively. This cognitive, discourse-level, and cultural adjustment makes passives particularly difficult in Japanese L2 learning. C-BLI is well suited to facilitate this sort of shift or restructuring. C-BLI designs the pedagogical environment to incorporate UB linguistic and cultural concepts, materializing and presenting them with SCOBAs, and guiding internalization via languaging (Swain et al., 2009). An appropriately designed C-BLI unit, thus, can help learners to learn and internalize how Japanese passives are used to express the experiencer-centered construal patterns of Japanese.

2.2. Applying UB Approaches to Japanese Grammar, Speaker’s Perspective, and Passives

The usage-based model of language, a core paradigm of cognitive linguistics, views grammar as a patterned system of symbolic structures that is, ultimately, a product of use (Bybee & Hopper, 2001; Bybee, 2006; Langacker, 2000). Language development emerges from exposure to recurrent exemplars, meaning that L1 acquisition is rooted in actual usage rather than in abstract grammatical rules (Tomasello, 2003). L2 learning is similarly item-based. Learners rely on previously experienced constructions while working to segment speech, map meanings, and discern combinatory patterns (Ellis, 2002). Many studies have examined this process in English and European languages (e.g., Ellis & Ferreira-Junior, 2009; Eskildsen, 2015); the present study extends this work to Japanese passives in an L2 context.

Construal—the framing of experience by shifting perspective and prominence—is a central concept of the UB model (Langacker, 2008a). Japanese and English differ in their preferred construals. This means that speakers of the two languages frame events in different ways. Japanese favors subjective construal: it is subjectivity-prominent (framing events from the perspective of interactants) and ego-oriented (defaulting to the speaker’s orientation) (S. Iwasaki, 1993; Ikegami, 2005, 2016). Subjective construal means that events are most often expressed from the experiencer’s perspective, without that perspective being directly stated (Ikegami, 2005, 2016). What is presumed to be understood remains unstated. Ikegami’s (2016) examples, shown in Examples (4a) and (4b), contrast Japanese and English utterances for the same event, that of being lost. In Example (4a), the Japanese speaker locates herself within the situation, so does not mention herself. In Example (4b), in contrast, the English speaker does not say ‘where is this’ but explicitly places herself into the scene. In Ikegami’s (2016, p. 305) words, ‘she undergoes a self-split mentally, herself stepping out of the situation and observing the situation from outside.’ English requires the speaker to encode the event in a more objective way. See Masuda et al. (2025, pp. 8–10) for more examples and discussion.

(4)	Being lost and asking the way (Ikegami, 2016: p. 305; glossing revised)
	a.	Koko wa doko desu ka?
		Here TOP where COP Q
		‘Where is this?’
	b.	Where am I?

Subjective construal tendencies shape various areas of Japanese grammar, including usage of motion verbs and benefactive, passive, and causative constructions (Masuda & Ohta, 2021; Masuda et al., 2025; Masuda & Ohta, forthcoming). For example, with motion verbs, the words iku and kuru are chosen based on whether movement is away from or toward the experiencer. A similar perspective-based distinction is found in the benefactive system, where verbs meaning ‘give’ (ageru and kureru) are differentiated based on the direction of giving relative to uchi, which comprises the speaker and/or the speaker’s in-group. Passives, which are our focus here, are most often selected to show impact on the experiencer or on someone the speaker has empathy with, whereas English typically prefers active sentences. Subjective construal has been widely discussed in Japanese and cognitive linguistics (Morita, 2016; Iori, 2022; Ikegami, 2005, 2016; Ikegami & Moriya, 2009); however, we have yet to find a Japanese language textbook that teaches this concept when introducing the passive in Japanese,6 and have not found any research, except our own, on the impact of teaching construal patterns to L2 learners. Because subjective construal conceptually underlies the usage of Japanese passives, which embody experiencer-centered, empathy-laden perspectives, understanding construal differences is crucial for Japanese L2 learners (Masuda et al., 2025, Masuda & Ohta, forthcoming). The UB model understands language acquisition as a bottom-up process, where rules emerge as learners use language, rather than being learned first and then applied deductively. For example, L1 learners do not first acquire an abstract rule for Japanese passives; instead, they use item-based passive forms available in their environment, gradually expanding their repertoire (Bybee, 2008). For primary language acquisition, early exposure to frequent, similar exemplars is especially effective before extending to greater variation (Bybee, 2008; Goldberg & Casenhiser, 2008). However, research on instructed SLA has found that it is most effective to incorporate both top-down and bottom-up processes, using explicit instruction in a rich L2 environment (Ellis, 2015; Tyler, 2012). Instructed SLA allows us to structure the learning environment to promote the learning of high-frequency passive verbs. Input-driven instruction without any explicit instruction is insufficient due to the lack of contact time for foreign language learners, since most foreign language learners lack extensive out-of-class language use opportunities. L1 transfer also works against just picking up complex structures that do not exist or are radically different from L1 structures. Students are unlikely to inductively learn underlying concepts, like subjective construal. Helping L2 learners to develop sensitivity to structures and their meanings that are substantially different from L1 thus involves systematic intervention.

2.3. How Are Japanese Passives Used, Acquired and Learned?

L1 acquisition studies show the prominence of passives that focus on the experiencer and often focus on negative experiences. Taguchi (2001) found that children first acquire passives as unanalyzed chunks tied to their daily experiences, e.g., okorareru ‘to be scolded.’ Mitsugi and Fukada (2022) found child-directed speech to contain a skewed set of exemplars (80% with negative affectedness) and frequent co-occurrence with shimau ‘put away,’ which adds a nuance of unwanted outcome when paired with a negative context—again centering the experiencer’s stance. Iori (2012) found that affective passives were about 14% more frequent than adversative passives, showing the importance of both and recommending early emphasis on affective passives in line with Kuno’s (1978) viewpoint hierarchy.

In both Japanese and English, passives allow the speaker/writer to avoid expressing the agent, so are useful when the agent is unknown. In Japanese, by expressing the impact on the experiencer, passives highlight the speaker/writer’s viewpoint (Koga, 2018; Sadanobu, 2022). Maeda’s (2011) analysis of film scripts found that 85% of the passives used in complex sentences highlighted the experiencer; about a quarter of these appeared in continuative -te clauses, with a discourse function of preserving subject continuity. Similarly, Mitsugi (2018) identified iwareru ‘to be told’ as a prototypical passive, emphasizing psychological impact (see also Takeda, 2005 for a similar finding).7 Related to the discourse function of passives, Minami (2021) compared Japanese L1 speakers with Japanese language learners, finding that Japanese L1 speakers used passives at story climaxes to sustain focus on the experiencer, whereas English-speaking Japanese language learners used active forms, shifting subjects depending on who was performing the action, which demonstrated the impact of English L1 transfer. Yabuki (2017) and Yabuki and Okuno (2020) likewise observed that learners underused passives in Japanese, relying instead on affective expressions (zannen, bikkuri, taihen) to show the impact of actions on the experiencer. Ōzeki (2023) compared naturalistic Japanese L2 learners with classroom learners, using data from the publicly available International Corpus of Japanese as a Second Language (I-JAS). She found that naturalistic learners acquired passives earlier than classroom learners, beginning with a limited set of verbs (such as iu ‘say’ and okoru/shikaru ‘scold’) and extending to -te form passives, such as iwarete. Naturalistic learners began using perspective-related passives earlier than classroom learners did. Some of these patterns, consistent with Maeda’s (2011) corpus findings, support a bottom-up, usage-based model of language acquisition (Bybee, 2006) and suggest that the language instruction experienced by the learners may not have provided them with the foundation they needed to use Japanese passives more naturally.

In sum, the appropriate use of Japanese passives requires L2 learners to have knowledge of forms and a strong understanding of how Japanese passives focus on the experiencer, highlight stance or affectedness, and sustain focus on the protagonist in discourse. Following Langacker (2008b, p. 83), learners should be introduced to prototypical uses, such as iu and shikaru/okoru, that occur commonly in Japanese passive usage. The next section discusses how we used C-BLI to engineer the classroom environment to promote the acquisition of passives, with a focus on their use to show impact on the experiencer, using concept-based language instruction.

2.4. Concept-Based Language Instruction, SCOBAs, and Languaging

Concept-based language instruction (C-BLI) is an approach to transforming classroom L2 instruction by foregrounding the linguistic and embedded cultural concepts essential for promoting L2 development. C-BLI is grounded in sociocultural theory (SCT). Originally called Cultural-Historical Psychology, SCT (Vygotsky, 1987) is a theory of human higher cognitive development that understands development from a holistic perspective that encompasses human language and culture, including thinking, speaking, and being in the world. The use of SCT as a framework for researching and understanding SLA has grown tremendously over the last thirty years, moving from descriptive (for example, Ohta, 2001) to interventionist (Masuda et al., 2025; Masuda, forthcoming; Swain et al., 2009; Kissling & Muthusamy, 2022; Lantolf & Tsai, 2018) research that works to promote SLA (Ohta, 2024). C-BLI is an interventionist approach to teaching/learning that emerged from Gal’perin’s (1992) systemic-theoretical instruction, a sociocultural approach to promoting children’s higher cognitive development via schooling. Applied to teaching foreign languages, C-BLI involves teaching/learning through concepts, to provide deeper understanding and foundations for further learning, rather than teaching simplified rules such as those common in foreign language textbooks. C-BLI is a multi-modal approach to instruction, meaning that it incorporates both language and visual components. A central component is the use of SCOBAs, which present visual materializations of concepts (Ohta, 2017). The role of SCOBAs is to accurately represent concepts so that they can function as cognitive tools for learners to use and, through their use, internalize. This study incorporates SCOBAs from Masuda and Ohta (forthcoming) to support instruction on Japanese passives.

Promoting internalization processes is a central goal of C-BLI, and this is done through a process of mediation, where learners use semiotic tools that become vehicles for their own thinking. Internalizing means developing self-regulation, or the ability to function without reliance on external tools. C-BLI curricula promote the internalization of concepts by engaging learners in verbal thinking, called languaging (Swain, 2006, p. 98), ‘the process of making meaning and shaping knowledge and experience through language.’ Languaging positions L2 learners as active agents in their conceptual development as they engage in collaborative and reflective meaning-making using SCOBAs; languaging about and with SCOBAs works to move abstract concepts from the external environment into the learner’s own thinking. Languaging is both a tool for students’ thinking/communicating and a mediational means for fostering self-regulation; when self-regulation is achieved, this means that learners have internalized the concepts taught and can thus function well without the support of SCOBAs. Prior studies have shown how languaging facilitates the acquisition of complex L2 grammatical structures such as a Japanese aspect markers (Tsujihara, 2022), Japanese polysemous particles (Masuda & Labarca, 2015, 2018), Spanish aspect (Kissling & Muthusamy, 2022), French voice (Swain et al., 2009), and English verbal collocations (Lantolf & Tsai, 2018).

Building on these theoretical foundations, the present study investigates how C-BLI supports classroom L2 learners’ acquisition of Japanese passives. Specifically, we examine how learners internalize the experiencer-centered construal that underlies natural passive usage. To capture this process, we analyzed learner production across multiple modalities, including story-writing and oral languaging. We assessed student development three times via an immediate post-test (post-test 1), a three-week-delayed post-test (post-test 2), and a six-month-delayed post-test (post-test 3), allowing us to trace both initial learning and the durability of the usage-based instructional intervention over time. Together, these measures provide a comprehensive view of how C-BLI mediational tools foster both usage-based conceptual knowledge and discourse-sensitive use of Japanese passives. Unlike traditional experimental designs that contrast control and treatment groups, this process-oriented study follows students’ developmental trajectories via analysis of their use of passives in writing and the conceptual knowledge expressed in their oral languaging. This approach enables us to capture conceptual development as it unfolds and better understand development over time, extending beyond the confines of a traditional academic term.

2.5. C-BLI of Japanese Passives

This section outlines our concept-based curriculum, which is based on a UB conceptual analysis of Japanese passive constructions. Before the passive unit, C-BLI was used to introduce the students to subjective and objective construals, motion verbs, and the benefactive construction, including the use of SCOBAs to teach semantic and grammatical concepts (see Masuda & Ohta, forthcoming). While passives can be classified in various ways, this study focuses on meaning and introduces three types of passives: the affective passive, adversative passive, and non-human passive. All passives consist of three elements:

• Experiencer: the person affected/impacted by the passive verb.
• Agent: the entity performing the action that affects/impacts the experiencer.
• Action: the event that happens to the experiencer, expressed by the verb.

In Japanese, the minimal unit of the passive construction is the passive verb itself; subjective construal of events in Japanese means that experiencers and agents are often omitted. The teacher introduced affective passives by dictating a short passage that focused on a protagonist’s experiences, expressed via passive sentences. The instructor showed the image of an ocean wave and had students listen to the sound of waves crashing to promote the association between the force of a wave and the meaning of Japanese passives. The teacher then explained the affective passive using SCOBA 1 (Figure 1), which superimposes a schematic visualization of the structure of the Japanese passive on the image of an ocean wave (for the purposes of this paper, we have romanized Japanese expressions on our SCOBAs). She explained SCOBA 1’s components by drawing students’ attention to the left side of the SCOBA; moving from left to right, the circle illustrates the source of impact generated by the agent, the arrow represents the direction of the verb’s action onto the experiencer, and the face icon represents the experiencer.

To show the relationships between the elements of the SCOBA, the teacher used gestures, modeling the perspective of the experiencer, with the passive verb crashing toward herself, prompting students to do this also, to show how a passive verb impacts the experiencer. The goal here was to guide students to experience the direction of a passive verb toward the experiencer.

To give students experience with a specific situation, the right side of SCOBA 1 illustrates a concrete context, a dog (the agent) barking at a woman (the experiencer). The dog looks scary, and the woman looks frightened. The face icon shows shock/surprise. This affective passive sentence can be translated into English as ‘The woman was barked at by the dog,’ but this does not convey the affectedness that is encoded into the Japanese passive sentence. Most likely, an English speaker would say ‘The dog barked at the woman,’ using an active verb, and adding an adjective (like ‘scary’) to describe the dog or a complement (like ‘frightening her’) to encode the negative impact. Without these, the English sentences, whether active or passive, have a neutral tone—we do not even know if the woman noticed the barking dog. In contrast, the passive affix used in Japanese, inu ni hoerareta ‘[Experiencer was] barked at by the dog’ conveys that the woman was affected by the barking. Here, that effect is negative. During instruction, we showed variations of SCOBA 1 showing a variety of sentences, including positive examples, such as Sensei ni sakubun o homerareta ‘[Experiencer’s] composition was praised by the teacher.’ We also included the verb iu ‘say,’ which is one of the most frequently passivized Japanese verbs. Passivized iu ‘say’ is often used in the continuative -te form, followed by positive information, as in tomodachi ni sugoi tte iwarete ureshikatta, ‘I was happy when my friend said “Amazing!”.’ Providing a variety of examples helps learners grasp the full range of passive usage, as they view the SCOBA for each example and use wave gestures to mime the impact on the experiencer.

Adversative passives are typically constructed with intransitive verbs, creating a structure that cannot be directly translated into English, which does not passivize intransitive verbs (in English we cannot say sentences like ‘I was cried by the baby’). Comparing the adversative passive with the affective passive in Japanese, Kikuchi and Masuda (2022) found that the former has a stronger nuance of harm; they analyzed verbs’ action-directedness (sashimuke), which is the extent to which the agent’s presence is implied in an action. We illustrate the adversative passive in SCOBA 2 (Figure 2), along with the sentence akachan ni nakareta. The illustrations show a crying baby (source of impact) and the negatively impacted caregiver. The active version of this sentence, akachan ga naita ‘The baby cried’ does not convey the negative impact on the caregiver, which the adversative passive encodes.

Finally, the unit also introduced the non-human passive, using SCOBA 3 (Figure 3). This is the easiest type of passive for English L1 students, since it is quite similar to English passive usage. Non-human passives are often used in written and formal discourse in both English and Japanese. In non-human passives, an invented or manufactured object serves as the subject, which is the focus of the sentence; the agent may be unknown or omitted if it is not important. In Japanese, this structure features non-human subjects who are well known or thematically significant (like the name of an inventor, author, or founder), so it sounds a little odd if ordinary subjects (like ‘my sister’ or ‘her friend’) are used. Our SCOBAs thus provide prototypical examples where the non-human subject would contribute meaningfully to the discourse topic. Here, the schematic SCOBA includes the non-human subject and the particle ni yotte, which is required to mark the source of impact if mentioned.8 On the right side of Figure 3, a specific example is given: ‘The light bulb was invented by Thomas Edison,’ and both an active and passive sentence were displayed.

In the classroom, SCOBAs 1–3 along with multiple examples were used to make the concepts visible and tangible. Along with the SCOBAs and explanations, authentic materials, such as Terrace House clips and their corresponding illustrated transcripts were also displayed, to show passive meanings in action. In-class tasks promoted internalization by having students apply the SCOBAs in pairs, working to construct passive sentences, acting out predicate actions, giving peer feedback on one another’s written sentences, and discussing experiential examples. By combining visual, auditory, and kinesthetic modalities, learners were guided to learn and internalize these concepts and to use them in spoken and written discourse, with oral languaging and written homework continuing this process.

3. The Study

To understand the impact of C-BLI passives instruction on learners’ L2 development of Japanese passives, we conducted a small qualitative longitudinal study in an intact fifth-semester Japanese language class.

3.1. Research Questions

This study addressed the following two research questions:

RQ1: How did a C-BLI approach to passives instruction support the participants’ overall understanding of passives?

RQ2: To what extent did the learners become able to deploy the concepts taught in the unit in their own writing?

To address RQ1, changes in conceptual understanding were examined over time via analysis of students’ oral languaging immediately after instruction, three weeks later, and six months later. To address RQ2, we examined the forms and functions of Japanese passives in personal narrative writing assignments.

3.2. Participants and Study Design

As mentioned earlier, this study is an extension of Masuda and Ohta (forthcoming), conducted in the same intact Japanese class.9

Participants: The twelve students enrolled in the first author’s fifth-semester Japanese course (designed for students with approximately 200 hours of prior classroom instruction) were invited to join the study; the students provided informed consent. Nine students had full attendance and completed all assignments included in the study.10 These nine students were invited to continue study procedures beyond the end of the semester. Five students agreed to participate. The students’ general proficiency in Japanese was ACTFL-OPI Intermediate-Mid.11 Students are identified by pseudonyms in this article. The students were from diverse cultural and linguistic backgrounds, including three English monolingual speakers (Isaac, Bill and Erica), one Chinese L2 English speaker (Alice), and one English native speaker with low conversational proficiency in Marathi from use of that language in his home growing up (Greg). Three students had been to Japan; Greg completed an eight-week internship in Tokyo, predominantly using English at work, and Isaac and Bill each spent a week in Japan as tourists; Alice and Erica had no prior experience visiting Japan. After completion of the Spring semester course, Bill, Erica and Isaac studied abroad for nine weeks in Oita, Japan. Alice returned home to China for the summer; Greg, the only graduate student in the study, spent the summer doing research on campus in the US.

Study design: The study design followed these five students through the course and for six months beyond course completion, collecting materials for the researchers to analyze to better understand their language development.

The course was primarily taught in Japanese, except for the C-BLI interactive lectures. C-BLI on Japanese passives was delivered through an interactive lecture, primarily in English, with Japanese incorporated. Students used English for oral languaging homework and Japanese for in-class and homework practice making/using Japanese passives. The curriculum is built on the concept of subjective construal and its impact on Japanese grammar, which was taught in the first week of the course. SCOBAs to teach subjective construal and an interactive lecture on the topic were incorporated (see Masuda et al., 2025, p. 11). As a part of the subjective construal unit, students conducted discourse analysis of a conversation in Japanese from the reality TV show Terrace House (Netflix) to identify first-person pronoun use/non-use and other features showing the subjective perspective in Japanese. Prior to the unit on passives, other C-BLI units were conducted on motion verbs and giving/receiving verbs. The passives unit began in the fourth week of the course; after the unit on passives, causatives were taught. Review exercises were incorporated throughout the course to consolidate understanding.

Table 1. C-BLI unit on Japanese passives.

	In-Class Activities and Homework Assignments (Study Components Are in All Caps)
4th week 1/30	PRE-TEST, a Homework 1: Assessment of learners’ prior knowledge of Japanese passives
5th week Day 1 of instruction 2/4	Class Activities 1 C-BLI interactive lecture on the basic concept of the passive (undergoer), including SCOBA 1, with many SCOBAs illustrating examples - Examples of passive sentences from Terrace House - SCOBA-supported explanation/practice of subject-focused passives, the particle ni for agents, and passive formation, using a wave gesture with both hands Practice: Apply the SCOBA to create passive sentences based on pictures Pair Work 1: Act out passive situations illustrated on SCOBAs, and then create passive sentences Pair Work 2: Based on a model dialogue, talk about personal experiences using passives Homework 2 (after class), practice forming passive verbs
Day 2 of instruction 2/6	Class Activities 2 Dictation and oral practice of short dialogues with passive sentences - Review of concepts and formation, with SCOBA 1 Interactive lecture on the adversative passive with SCOBA 2; explanation that passives can exist without active counterparts and their discourse function Introduction/practice of inanimate-subject passives, with SCOBA 3 Pair Work 1: Create passive sentences using cues and discuss troublesome roommates, supported by SCOBAs 1 and 2 Listening: Teacher’s horrible trip to New York IMMEDIATE POST-TEST (POST-TEST 1): Homework 3 (after class), writing simple passive sentences ORAL LANGUAGING (audio or video-recorded; see assignment instructions in the Appendix A) including explanation and analysis of passive sentences
6th week Day 3 of instruction 2/11	Class Activities 3 Review passives through listening practice Introduction and practice of other grammar items Homework 4 (after class), part of the immediate post-test, post-test 1 NARRATIVE PARAGRAPH WRITING, using passives, about a troubling or challenging experience Illustrated dialogues with blanks: Fill in the blanks with passive sentence elements
7th week Day 4 of instruction	Class Activities 4 Listening to and reading lyrics of the song ‘Love Letter,’ containing passive sentences Peer feedback on their ‘troubling experiences’ compositions and how passives were used Other unrelated activities (e.g., kanji, listening) Review sheet for the unit test
8th week	Unit Test: Includes written languaging tasks targeting passive constructions
10th week 3/12 and 3/14	POST-TEST 2, Homework 5 (3 weeks after the passives unit) ORAL LANGUAGING (audio and/or video-recorded) including explanation and analysis of passive sentences NARRATIVE PARAGRAPH WRITING, using passives, about a bad dream
Mid-August	POST-TEST 3 (6 months after the passives unit) ORAL LANGUAGING (video-recorded) including explanation and analysis of passive sentences, recorded on Zoom Follow-up interviews, recorded on Zoom NARRATIVE PARAGRAPH WRITING, using passives, about a bad summer day

^a Pre-test questions regarding passive knowledge were Q1: What do you know about the passive voice? Please write a few examples of passive sentences in English; Q2: What are some situations where you would use the passive?; Q3: Do you know anything about passives in Japanese? Please explain what you know.; Q4: Do you speak any other languages? How is the passive in that language similar to or different from English?

shows the study design and how it was embedded into the C-BLI unit on passives. The pre-test, post-tests, oral languaging, and student writing components relevant to the present study are noted in all caps.

3.3. Data Analysis Methodology

Our collection of longitudinal data allowed for an analysis focused on student development over time, integrating multiple modalities to provide a comprehensive view of cognitive and discourse-linguistic development of learners’ Japanese passives. The methods used to analyze the data collected are as follows.

Oral languaging: To address RQ1, on the development of learners’ conceptual understanding of passives, we analyzed audio data from the three post-test oral languaging sessions. For post-tests 1 and 2, the students audio- or video-recorded their own oral languaging, submitting it online (post-tests 1 and 2). Post-test 3 was conducted after a Zoom meeting with the teacher.12 Since the students were already present on Zoom, they were asked to perform the oral languaging during the Zoom meeting, after the teacher left the room. To investigate students’ conceptual development, the data were collaboratively transcribed and coded. Coding categories consisted of affective passive, adversative passive, inanimate subject passive, speaker perspective, subject omission, and grammatical particles. Students also used the terms direct/indirect passive, so we coded for this separately, in order to clearly identify what students said in their languaging. The first author and a research assistant,13 who was trained to identify the items to be coded, conducted this analysis. Discrepancies in coding were resolved through discussion.

To understand students’ conceptual development, we considered the frequency of the concepts coded. We quantified mention of concepts by each student and across students to see how student mentions changed over time and to compare students to one another.

Writing: Students’ written narratives, collected from the three post-tests (see Table 1 for writing topics) were analyzed, focusing on the use of passives. For all three topics, students were permitted to include fictional elements in their narratives. The type of verb, semantic function, type of predicate, obligatory contexts, correct use, misuse, partial misuse, over-use and non-use were coded by the two authors, with an interrater reliability of 86%. Discrepancies were resolved via discussion. TLU calculations followed Pica’s (1983/2008) formula. Analysis of these data were used to address RQ2, on development of the use of passives in student writing.

4. Results and Discussion

4.1. Conceptual Development of Japanese Passives (RQ1)

RQ1 asks how the C-BLI approach to passives instruction supported the participants’ overall understanding of passives. The students began with little to no understanding of passives; through analysis of the data, all of the students’ oral languaging showed growth in their conceptual understanding of the passives concepts taught via C-BLI. Pre-test findings showed no prior knowledge of passives for Alice, Bill and Erica; Greg and Isaac had little previous knowledge. Greg, who had spent a couple of months in Japan prior to this study, knew that passives could be used to express negative experiences. Isaac understood the general shape of the passive affix but listed two incorrect examples of what he thought were passivized verbs.

After instruction, all students showed conceptual growth. As a group, there were fluctuations over time.

Table 2. Mentions of passive-related concepts on post-tests 1–3.

	AfP	Mor	ni	AdvP	SP	OmSub	DpIp	Inan	Total
Post-test 1	4	4	5	2	2	0	0	0	17
Post-test 2	4	3	5	3	4	3	1	3	26
Post-test 3	5	5	3	3	2	0	3	1	22

Key: AfP (affective passive), Mor (morphological formation of the passive), ni (ni indicating the source/agent), AdvP (adversative passive), SP (speaker’s perspective), OmSub (omission of subjects), DpIp (affective and/or adversative passive), Inan (inanimate subjective passive).

displays student mentions of passives in their oral languaging. The most mentioned were the affective passive and the morphology of the passive. The number of students mentioning that passives relate to the speaker’s perspective (SP in Table 2) increased from two to four students from post-tests 1 to 2, then decreasing back to two students on post-test 3. No students mentioned inanimate passives (Inan) on post-test 1, but one did so on post-test 2, increasing to three students on post-test 3. All five of the students mentioned use of the particle ni on post-tests 1 and 2, decreasing to three of the students on post-test 3. It is doubtful that they forgot about this essential particle; however, they did not mention it.

Mentions of passives concepts increased dramatically from the immediate post-test (17 mentions on post-test 1) to post-test 2 (26 mentions). Post-test 3 also shows an increase over post-test 1 (22 mentions), though somewhat lower than post-test 2. Students did not talk about all of the concepts immediately. Some topics emerged only after a delay, particularly among students who spent the summer in Japan. We might understand this delayed surfacing of mentions to relate to an incubation period in second language acquisition (cf. Gass, 1997), as knowledge becomes observable over time.

There was variability across students in the concepts they mentioned (see

Table 3. Mentions of passive-related concepts, by student.

Name	Post-Test 1	Post-Test 2	Post-Test 3
Alice	AfP, Mor, ni	AfP, Mor, ni, Inan	Mor, AdvP
Bill	AfP, Mor, ni, AdvP, SP	AfP, ni, AdvP, SP	AfP, Mor, AdvP, SP, DpIp
Erica	AfP, Mor, ni, SP	AfP, Mor, ni, AdvP, SP, OmSub, DpIp, Inan	AfP, Mor, ni, DpIp
Greg	ni, OmSub, Inan	ni, SP, OmSub, Inan	AfP, Mor, ni, SP, Inan
Isaac	AfP, Mor, ni	AfP, Mor, ni, AdvP, SP, OmSub	AfP, Mor, ni, AdvP, DpIp

Key: AfP (affective passive), Mor (morphological formation of the passive), ni (ni indicating the source/agent), AdvP (adversative passive), SP (speaker’s perspective), OmSub (omission of subjects), DpIp (direct and/or indirect passive), Inan (inanimate subjective passive).

). Four of the five learners (all except Bill) maintained or deepened their understanding between post-tests 1 and 2. Those who studied in Japan (Bill, Erica, and Isaac) sustained their semantic understanding through post-test 3. Average conceptual categories mentioned rose from 4.3 to 5.5 across post-tests 1 and 2. Erica, new to the passive, cited all eight categories on post-test 2. The analysis supports an understanding of C-BLI as fostering both breadth and depth of conceptual development. From a usage-based perspective, learners had not merely memorized forms, but went through a process of building form–meaning mappings through conceptual framing, creating a foundation for a flexible and natural use of Japanese passives.

Though we deliberately included gestures as part of instruction, because we did not plan to study gestures we did not record the teacher’s gestures or ask students to video-record their oral languaging. Thus, we cannot provide a detailed analysis of the impact of pedagogical gestures. However, during post-test 3, which was on Zoom, Isaac sat far enough from the camera for us to partially see his gestures. He used ocean wave gestures when speaking about how the passive reverses direction of the verb’s impact, as shown in Figure 4. As mentioned earlier, the SCOBAs had arrows showing the passive verb’s impact on the experiencer, which the instructor taught with accompanying gestures. After mentioning ‘the image of the wave,’ he used both hands to produce wave-like gestures (see Figure 4), adding ‘Yeah, I think the image of the wave indicating the direction of action is really helpful in a sense.’ Isaac’s gestures support previous research on the value of pedagogical gestures (Stam, 2018; Stam et al., 2023).

4.2. Development of Passives in Story-Writing (RQ2)

Table 4. The types and tokens of morphemes used in story-writing.

	Types	Tokens	Token Average
Post-test 1	181	527	105.4
Post-test 2	237	747	149.4
Post-test 3	240	818	163.6

presents the overall types and tokens of morphemes used in the story-writing task, calculated following Co-Chu (Yamamoto et al., 2017). A ‘type’ is a unique instance of use; thus, this calculation shows the number of different morphemes used. A ‘token’ is an instance of use, including repetitions of the same type; thus, this calculation shows the total number of morphemes used. Students tended to write more on each subsequent post-test, as shown in the increasing number of tokens. The number of types also increased over time.

Table 5. The number of passives used in story-writing.

	Types	Tokens	Token Average
Post-test 1	9	13	2.6
Post-test 2	12	15	3.0
Post-test 3	18	21	4.2

provides type and token calculations for the passive verbs that the students wrote on the post-tests. The students produced more passives over time, as noted by the increase in tokens. There were 49 instances (tokens) of 39 different verbs (types); students used 13 passives (average 2.6) on post-test 1, 15 passives (average 3.0) on post-test 2, and 21 passives (average 4.2) on post-test 3, a substantial increase across the three post-tests. The types (different passive verbs used) doubled over the three post-tests.

Type–Token frequency of passive verbs:

We also calculated type–token ratios (TTRs) of passive verbs. This calculation measures lexical variety (Granger & Wynne, 2000), determined by dividing the number of types by the number of tokens, with a higher number showing more lexical variety. The five learners’ TTR increased from 0.69 on post-test 1, to 0.80 on post-test 2, to 0.86 on post-test 3; this is a modest increase in the variety of verbs used over time. For comparison, the TTR in a corpus study of L2 learners’ conversational passives was 0.41 (Mitsugi, 2018).

While TTR measures lexical variety, it does not capture distributional patterns. To understand the verbs used on each post-test, we calculated frequencies of the types used, as shown in

Table 6. Passive verbs used on the post-tests in students’ story-writing.

Post-Test 1		Post-Test 2		Post-Test 3
Verb	Freq.	Verb	Freq.	Verb	Freq.
iu ‘say’	3	nusumu ‘steal’	2	iu ‘say’	3
taberu ‘eat’	2	suru ‘do’	2	nusumu ‘steal’	2
torikesu ‘cancel’	2	furu ‘rain’	2	suru ‘do’	1
iku ‘go’	1	ou ‘chase’	1	iku ‘go’	1
nusumu ‘steal’	1	hiku ‘play an instrument’	1	shimeru ‘shut’	1
oku ‘put’	1	kuru ‘come’	1	taberu ‘eat’	1
yaku ‘burn’	1	utau ‘sing’	1	nomu ‘drink’	1
yogosu ‘soil’	1	bikkurisuru ‘be surprised’	1	ageru ‘give’	1
suru ‘do’	1	ijimeru ‘burry’	1	bikkurisuru ‘be surprised’	1
		ochiru ‘drop’	1	komaru ‘be troubled’	1
		hotte oku ’leave’	1	kau ‘buy’	1
		tasukeru ‘help’	1	kawaru ‘change’	1
				owaru ‘end’	1
				matsu ‘wait’	1
				yameru ‘quit’	1
				okureru ‘delay’	1
				wasureru ‘forget’	1
				otosu ‘drop’	1

. The stair-step design of the table shows the increase in types over time. On post-test 1, iu ‘say’ was used three times, while taberu ‘eat’ and torikesu ‘cancel’ each appeared twice, and six other verbs appeared once each. On post-test 2, iu ‘say’ was not used; nusumu ‘steal,’ furu ‘rain,’ and suru ‘do’ each appeared twice, and nine other verbs appeared once each. Lexical variety grew further on post-test 3, with iu occurring three times, nusumu twice, and sixteen other verbs used once each. While with this small data pool it is difficult to generalize, iu ‘say’ and nusumu ‘steal’ were the most frequently used verbs. The growing lexical variety shows students’ language development over the study period. Rather than a reduction in passives use over time, students’ usage of passives and lexical variety increased over time.

As mentioned earlier, previous corpus-based studies have found iu (passivized as iwareru ‘be told’) to be the most frequently passivized verb, among L2 learners (Mitsugi, 2018; Ōzeki, 2023) as well as L1 speakers (Mitsugi, 2018; Takeda, 2005; Nakamata, 2014); our small sample also demonstrated this finding, with samples of student usage shown in Example (5):

(5)	Student uses of iu ‘say’ in the passive
	a.	Ano yoru rūmumeito ni ‘warui’ to iwaremashita. (Isaac, post-test 1)
		‘That evening my room-mate said “sorry” to me.’
	b.	Bāsu ga sanjikan okureru to iwaremashita. (Greg, post-test 3)
		‘They told me the bus would be three hours late.’
	c.	*Hayai yoteisuru yoyaku wa ichigatsu desu to iwaremashita.14 (Erica, post-test 3)
		‘They told me that the earliest appointment I could schedule is January.’

This pattern reflects a broader tendency for Japanese passives to involve psychological and cognitive-experience verbs (Mitsugi, 2018, p. 119), unlike English, where prototypical passives typically involve transitive action verbs. Tanno (2010), in his study of how L2 learners acquire hearsay and quotation structures, observed that iu passives are often used to frame quotations, allowing the listener to identify the source of speech in the absence of an explicit subject. Here, we see these participants using passives very naturally. These findings highlight the central role of speaker perspective and agentivity in Japanese passive use and its development.

Semantic functions of the passive verbs: Ōzeki (2023) divides passives with animate subjects into four categories: negative affectedness passive (hereafter, negative), benefactive passive (hereafter, benefactive), perspective-related passive (hereafter, perspective), and backgrounding of the agent passive (hereafter, backgrounding). Ōzeki classifies passives with an inanimate subject as inanimate. The classroom learners in Ōzeki’s study tended to use negative passives most frequently up until the intermediate level. With increasing proficiency, L2 learners began using perspective and backgrounding passives. She compared classroom learner data with data from naturalistic learners, who used perspective passives earlier and frequently employed the widely used passive verb iwareru ‘be told.’ This pattern reflects the instructional order of typical Japanese language textbooks, which generally emphasize negative passives and show a negative impact on the experiencer.

Table 7. Passives used in post-test story-writing, by semantic function, showing percentage and number.

	Negative	Benefactive	Viewpoint	Background	Non-Human	Total
Post-test 1	75.0 (10)	0	25.0 (3)	0	0	100 (13)
Post-test 2	73.3 (11)	6.7 (1)	6.7 (1)	0	13.3 (2)	100 (15)
Post-test 3	71.4 (15)	0	14.3 (3)	9.5 (2)	4.8 (1)	100 (21)

The numbers in parentheses indicate passive occurrences.

displays data from our participants, showing them to follow the same pattern as Ōzeki’s classroom learners.15 Table 7 displays the percentages and number of negative, benefactive, viewpoint, background and non-human passives used in student writing in post-tests 1–3, along with, in the far right column, the percentage of these usages that were adversative passives.

It is not surprising that the students used so many negative passives (over 70% of all usages on the three post-tests), because they were asked to write about real or imagined negative experiences, and the curriculum spent relatively more time on examples expressing a negative impact on the experiencer, which was a new concept for students. Even so, they also used benefactive and viewpoint passives, which are perspective passives, albeit infrequently. Benefactive, background, and non-human passives also appeared on one or more of the later post-tests, as shown in Example (6):

(6)	Students’ use of benefactive, viewpoint, and non-human passives
	a.	Uchūjin ni …tasukeraremashita. (benefactive passive, Greg, post-test 2)
		‘(I) was helped by an alien.’
	b.	*Byōin ni ‘hayai yotei suru yoyaku wa ichigatsu desu’ to iwaremashita. (viewpoint, Erica, post-test 3)
		‘I was told by the hospital “January is an early appointment time.”’
	c.	*Shawā o shimerarete arimashita. (non-human passive, Bill, post-test 3)
		‘The showers were closed.’

Table 7 also shows an increase in adversative passives across the three post-tests; these generally lack English equivalents. These include Chichi ni aisukurīmu o taberaremashita ‘My father ate my ice cream’ by Bill on post-test 1, and *Ame ga furareta ‘It rained on me’ by Erica on post-test 2, which was introduced and practiced in class in the form of ame ni furareta. Sentences like this were practiced in class, contributing to the occurrence of adversative passives. On post-test 3, learners’ use of adversative passives exceeded 95% of all passives used.

Predicate types of the passive verbs: The participants used passives in a variety of predicate types. Following Ōzeki (2023), learners’ passive verbs were classified into the following categories: -masu/-mashita (sentence-final), -te imasu/-te imashita (aspect), -te shimau/-chau (regret), -te (showing the reason), and other. Ōzeki (2023) found that L2 learners first tended to use passives with sentence-final forms, only later using the -te form to show reasons, as in Shōkai sarete yūmei ni natta sōdesu ne ‘It was introduced, so became famous’ (Ōzeki, 2023, p. 110). In her data, about 20% of the passive usages of English L1 classroom learners were combined with the -te form. By contrast, in her naturalistic acquisition data, this predicate type appeared more often and from the earliest stages of passive use. By the late-intermediate level, over 62% of the naturalistic learners’ passives were in -te form predicates, with the most frequent expressions involving the passive verb iwarete ‘being told.’

The predicate types used by our study participants are shown in

Table 8. Passive predicate types in story-writing on post-tests 1–3, showing percentage and number.

	Sentence-Final	Te-iru	Te-shimau	Te-form	Other	Total
Post-test 1	61.5 (8)	0	7.7 (1)	15.4 (2)	15.4 (2)	100 (13)
Post-test 2	53.3 (8)	0	0	20.0 (3)	26.7 (4)	100 (15)
Post-test 3	66.7 (14)	0	0	19.0 (4)	14.3 (3)	100 (21)

The numbers in parentheses indicate passive occurrences.

. The proportion of -te forms increased from 15.4% to about 19–20%. Students did not produce te-iru predicates. One student used -te shimau, shown in example 7a. Students also used predicate types classified as ‘other,’ such as using passives in subordinate clauses expressing reasons, with the conjunctions node or kara. Examples (7b) and (7c) show students using a passive with the conjunction kedo ‘but’ and another using the passive with te-aru, which indicates an action done in advance. In Example (7), the participants’ passives appropriately describe situations where the actions caused them some difficulty.

(7)	Student passives describing difficult situations
	a.	Rūmumeito ni bentō o taberarete shimai mashita (te-shimau, Isaac, post-test 1)
		‘My roommate ate my lunch.’
	b.	Tomodachi ni hotte okareta kedo… (other, Greg, post-test 2)
		‘My friend left me by myself.’
	c.	*Shawā o shimerarete arimashita. (other, Bill, post-test 3)
		‘The showers were closed.’

Target-like use (TLU) of passive verbs in student story-writing: The correctness of passives was analyzed via TLU calculations (Pica, 1983/2008), as shown in

Table 9. TLU scores for passives in participants’ written stories.

	Alice	Bill	Erica	Greg	Isaac	Average
Post-test 1	17 (3)	100 (1)	50 (1)	66 (3)	80 (5)	62.6
Post-test 2	50 (2)	37.5 (4)	87.5 (4)	50 (4)	75 (2)	60.0
Post-test 3	33 (3)	33 (3)	83 (3)	43.8 (8)	100 (4)	58.6

The numbers in parentheses indicate passive occurrences.

. This is a measure of the correctness of morphemes, analyzed by considering correct use, misuse, and underuse in the given context; taken alone, a TLU score provides only a partial picture of L2 development.

From post-test 1 to 3, the average TLU score declined slightly, from 62.6 on post-test 1 to 60.0 on post-test 2, and to 58.6 on post-test 3. However, there was considerable variation across individuals, which is to be expected, as individuals develop L2 skills at different rates; these learners also had widely varying opportunities to use Japanese during the five-month period following course completion. Here, we provide a description of students’ errors in their written stories, since the small number of usages of the passive results in TLU scores do not effectively capture the students’ development. These descriptions provide a better view of student development, complementing our other measures, which provide positive evidence of development.

• Alice sometimes overused passives where te-shimau would have been appropriate. She expressed surprise by saying bikkuri sareta ‘was surprised,’ perhaps a direct translation from English, where, in Japanese, using te-shimau/chau is natural, as in bikkuri shichatta ‘was (completely) surprised.’ Her TLU score improved on post-test 2 but declined on post-test 3.
• Bill used just one passive verb on post-test 1, and it was correct, which resulted in a TLU score of 100. On subsequent post-tests, he attempted multiple new verbs. His uses were often inaccurate, but he showed a strong conceptual grasp of passives for expressing a negative impact on the writer. He creatively described being hit by bird droppings *Tori kara kami ya atama ni purezento o otosaremashita, ‘From the bird, presents were dropped on my hair and head,’ and also described the shower being turned off on him *Shawā o shimerarete arimashita, ‘The showers were turned off,’ and missing a quiz because he arrived late *“Poppu kuizu ga ageraremashita,” ‘A pop-quiz was given.’ These clearly convey the negative impact of these actions on him. Like Alice, he also over-used passive in a context where the -te shimau/chau predicate would have been correct.
• Greg used passives a lot on post-tests 2 and 3, making many errors, and also overused passives, sometimes producing a passive where a causative-passive was required (e.g., on post-test 3, matareta instead of matasareta ‘was made to wait’).
• Erica developed more slowly than the others, moving from just one use on post-test 1 to four uses on post-test 2 and three on post-test 3. On post-test 2, she used a variety of verbs, including writing *Ame ga furaremashita, ‘It rained,’ hana ni aruite koraremashita, hana ni ijimeraremashita ‘The monster flower came to me and he bullied me.’ On post-test 3, she used the passive correctly twice, but once she passivized the wrong verb: *…renraku shite komararemashita instead of … renraku sarete komarimashita, ‘I was contacted and it was annoying.’
• Isaac consistently produced correct passives but generally relied on memorized chunks (expressions used in class and in the textbook) rather than using language more creatively. This strategy resulted in generally high TLU scores compared to the other students. Despite the lack of risk-taking in his writing, his oral languaging showed mastery of passives-related concepts, including via his gestures, as was discussed in Section 4.1.

As indicated in the discussion, the slight decline in average TLU scores may mask their progress; the data suggest that students are following typical U-shaped developmental trajectories. Correctness is important, but the increase in lexical variety and creativity of language use that involves extending passives to new verbs and contexts shows students’ language learning progress. The participants’ active engagement with meaning and form, demonstrated through their growth in using passives with a variety of verbs and contexts, is consistent with the incubation period in L2 development, described by Gass (1997); in this process, learners experiment with new forms before achieving stable, accurate use. Also, from a usage-based perspective, the variability students show in their story-writing and oral languaging is ‘never noise and instead always holds developmental relevance’ (Tyler & Ortega, 2018, p. 317).

These data are a good illustration of the non-linearity of L2 development which, through a sociocultural lens, follows a zig-zag pattern of growth and seeming regression, as students struggle to express themselves in the L2 in new contexts. The participants’ writing provides rich evidence of their use of target forms to create meaning.

5. Conclusions

In sum, our findings show evidence of student development linked to a C-BLI unit, designed using usage-based linguistics. Learners developed both conceptually and linguistically, deploying the concept taught to support their written storytelling and demonstrating high aspectual accuracy, showing how conceptual understanding supported their meaningful production.

Here, we present how the classroom linguistic environment, created using usage-based linguistics and C-BLI, resulted in students’ development as they learned Japanese passives for the first time. Our instruction focused on the core grammatical concept of how passivized verbs impact the subject or, to use a linguistic term, the concept of affectedness. By materializing core concepts into SCOBAs, used for teaching explicitly in interactive lectures that incorporated salient pedagogical gestures, promoting internalization via oral languaging tasks, and providing ample Japanese language practice in and outside of class, learners were guided step by step toward the central discourse-level concept of the passive. Instruction emphasized the core meaning of ‘undergoing psychological or physical impact,’ visually conveyed through the metaphor of a ‘wave’ and highlighting the perspective of the affected participant.

Analyses of oral languaging showed that the students could explain the core concepts; for the student whose gestural space we could clearly observe, our findings also suggest the helpfulness of the ‘wave’ gesture in conveying the concept of passive-verb-actions impacting the speaker, as evidenced by the student’s thinking-for-speaking. Analyses of the participants’ written stories showed both individual differences and evidenced solidification of learners’ conceptual understanding, which became more systematic over time, even though accuracy suffered, as their usage of and creativity with passives increased, often reducing TLU scores. Nevertheless, prototypical uses of the passive, such as iwareru ‘to be told’ were observed, and our findings via the 6-month delayed post-test 3 portend development over time in the usage of passive for negative/perspective functions as well as complex morphological forms (e.g., te-form).

Studying students’ developmental process in intact classrooms is an important and challenging area of research. It is important because it shows us how L2 development can be shaped by classroom instruction and, in this case, how usage-based linguistics concepts, materialized and taught using C-BLI, impact learners’ development of Japanese passives. Classes in less-commonly-taught languages are often small; there may be just one class per level, limiting the potential number of available participants and resulting in a small study size that precludes the use of inferential statistics or claims of generalizability. The small number of students also eliminates the option of including a comparison group or in using the experimental method. On the other hand, our intention was to carefully look at the development of particular learners, in order to capture their developmental processes, using an intact class to provide high validity. In future research, we plan to consider how Dynamic Assessment (Lantolf & Poehner, 2014, Chap. 8) might provide more individualized support and allow us to better understand students’ thinking as they work to apply their knowledge of passives in the process of writing revision. Meanwhile, the present study highlights how usage-based and C-BLI approaches complement one another: usage-based learning grounds instruction in frequency and authentic use, while C-BLI provides conceptual mediation for internalization. Together, they offer a powerful model for teaching difficult areas of Japanese grammar, where authentic usage and conceptual understanding can best develop hand in hand.