1.1 Vocabulary and kanji learning

Among the studies yielded by databases search with keywords such as ‘vocabulary learning’, ‘lexical development’, and ‘L2 Japanese’, over 80% are concerned with a range of challenges associated with learning kanji. This fact indicates that learning a large number of kanji characters and words (i.e., single character words and multiple-character compounding words) has been a major concern among L2 Japanese researchers and educators. Since kanji words constitute a crucial part of Japanese written and professional vocabulary, the acquisition of a sufficient working knowledge of characters and kanji words is vital in skilled reading and professional communication (Everson, Reference Everson2011; Mori, Reference Mori2014). The learning of kanji involves not only the study of orthographic features of characters but also the buildup of essential vocabulary that is commonly used in professional, academic, and even daily settings. It also involves the acquisition of different types of lexical knowledge including its semantic and phonetic values, orthographic features, compositional structures, and grammatical functions. In addition, L2 word recognition and processing operate a qualitatively different mechanism from that of L1 word identification, requiring increased cognitive resources for orthographic, semantic, and phonological retrieval (e.g., Fukuta, Reference Fukuta2013; Koyama, Stein, Stoodley, & Hansen, Reference Koyama, Stein, Stoodley and Hansen2013; Nakayama, Ida, & Lupker, Reference Nakayama, Ida and Lupker2016). Consequently, post-2010 studies continued to investigate a range of issues in learning kanji from various perspectives including crosslinguistic influence (Matsumoto, Reference Matsumoto2013; Hagiwara, Reference Hagiwara2016), the role of L2 proficiency (Toyoda & McNamara, Reference Toyoda and McNamara2011; Matsumoto, Reference Matsumoto2013), linguistic factors affecting L2 kanji recognition and production (Fei & Matsumi, Reference Fei and Matsumi2013; Toyoda, Firdaus, & Kano, Reference Toyoda, Firdaus and Kano2013; Taniguchi, Reference Taniguchi2017), learning strategies and attitudes (Soemer & Schwan, Reference Soemer and Schwan2012; Rose, Reference Rose2013; Thomas, Reference Thomas2013, Reference Thomas2014, Reference Thomas2015; Tsukada, Reference Tsukada2012; Kubota, Reference Kubota2017), and instructional interventions (Ulambayar, Reference Ulambayar2013).

Pre-2010 studies had revealed that L2 Japanese learners with different L1 orthographic backgrounds demonstrate different degrees of phonological and visual reliance in processing kanji words (see Mori, Reference Mori2014 for a summary). L2 learners whose L1 is alphabetic (i.e., a language employing a sound-based orthography, e.g., English), for instance, exhibit more phonological reliance and less visual reliance in L2 word recognition than those whose L1 is logographic (i.e., a language using a morpheme-based orthography, e.g., Chinese). However, it remained unclear whether L1-based orthographic processing strategies would persist or diminish with increased L2 proficiency. To answer this question, Matsumoto (Reference Matsumoto2013) compared the lexical judgment and reading skills of three groups of L2 Japanese college students (i.e., beginning-level students whose L1 was alphabetic [English], beginning-level students whose L1 was logographic [Chinese], and intermediate-level students whose L1 was English). It was not surprising that the Chinese group, who possessed sufficient character knowledge, outperformed the other two groups on the lexical judgment task. However, the higher proficiency L1 alphabetic group exhibited more visual reliance than its lower proficiency counterpart, suggesting that L1-based orthographic processing strategies could adjust in accordance with L2-specific orthographic features. Consistent results were reported by Toyoda and McNamara (Reference Toyoda and McNamara2011) as well.

Hagiwara (Reference Hagiwara2016), however, challenged the effect of L1-based orthographic processing strategies on L2 kanji learning by reporting that L2 Japanese learners from an L1 alphabetic background, who were expected to use sound-based strategies, demonstrated the ability to learn novel kanji successfully regardless of the presence of phonological mediation. In her study, 70 college students studying in Japan whose L1 was alphabetic were divided into three learning conditions (i.e., pronunciation-hearing and read aloud, pronunciation-hearing only, and no pronunciation-hearing) and learned two types of novel kanji (i.e., characters that contained visually emphasized phonetic components and ones without emphasis). No statistically significant difference was found across the groups in either meaning recall or character identification. While this finding might suggest no major role of sound information in semantic and/or orthographic learning, the effectiveness of the sound-based strategies on sound learning remained unclear since the participants were not tested on the pronunciation of the target words. The finding that the participants paid more attention to the phonetic components in the pronunciation-hearing conditions than in the no pronunciation-hearing condition indeed suggested that sound-based strategies raised awareness of phonetic clues available in visual representations.

In fact, recent years have seen an increasing interest in exploring an effective use of phonetic information in kanji learning, as the majority of joyo (everyday use) kanji comprises semantic-phonetic composite characters (keisei-moji). Toyoda et al. (Reference Toyoda, Firdaus and Kano2013), for instance, identified 119 potentially useful phonetic components for L2 Japanese students, mainly for those from alphabetic backgrounds, based on the components’ various attributes including frequency, likelihood that L2 readers encounter, the number of pronunciations, phonetic reliability, and productivity. While it still needs to be empirically validated whether or not L2 Japanese learners are able to take advantage of the availability of phonetic components, there has been evidence suggesting that L2 learners use sound information in L2 kanji learning. Fei and Matsumi (Reference Fei and Matsumi2013), for instance, investigated Chinese-speaking L2 Japanese learners’ processing of auditory-presented kanji words, which differed in phonetic or orthographic similarity to Chinese counterparts, while listening to Japanese sentences providing contextually strong versus less strong clues for the target words. The interactional effect of phonetic and orthographic similarities was found only for the contextually strong sentences, but not for the contextually loose ones, suggesting that use of phonetic and orthographic information played a role in processing auditory-presented kanji words in relation to the degree of contextual constraints.

Coping with visual complexity is another area of interest as it is a major inhibitory factor for character reproduction, particularly for non-kanji background learners (e.g., Taniguchi, Reference Taniguchi2017). Strategies under investigation include mnemonics (i.e., memory-enhancing strategies) (Toyoda & McNamara, Reference Toyoda and McNamara2011; Soemer & Schwan, Reference Soemer and Schwan2012; Rose, Reference Rose2013) and rote learning such as repeated writing (Tsukada, Reference Tsukada2012) and air writing (i.e., spontaneous production of highly articulated movements of the fingertips or hands while tracing out abstract representations of kanji, or character components) (Thomas, Reference Thomas2013, Reference Thomas2014, Reference Thomas2015). Studies on mnemonics through the use of personal encoding techniques, or mental images of the items to be learned, for instance, have not reached researchers’ consensus on their effectiveness. Mnemonic devices, including keywords, visual images, and phonological resemblance to the target items, and memory aids that can either be formally pre-established or idiosyncratically self-generated, have been found to be useful when being applied in a meaningful way. Soemer and Schwan (Reference Soemer and Schwan2012), for instance, demonstrated that Sino-Japanese characters and their meanings were learned better with either static pictures or animated morphs as visual mnemonics than in a non-mnemonic condition. Between the two types of visual mnemonics, static pictures showed a superiority over animated morphs, but the advantage of static pictures disappeared when the length of presentation time of characters and pictures was set constant. This suggested that the associability of characters’ visual shapes with their meanings, rather than types of encoding mnemonics, affect success in semantic learning. Rose (Reference Rose2013), however, cautioned that an overreliance on mnemonics could lead to some negative outcomes in English-speaking college students’ kanji learning, pointing out such drawbacks as loss of the meaning of a character in overly complex mnemonics and difficulty in generating sufficient sound associations.

Repeated writing, which is commonly used to learn to write characters, was found to be an effective strategy for L2 learners when being used in meaningful contexts (Tsukada, Reference Tsukada2012). Air writing is also often used by adult learners of Japanese residing in Japan, just as native readers do, but its effect was unclear. Thomas (Reference Thomas2014, Reference Thomas2015) investigated the effectiveness of air writing for L2 Japanese learners in memorizing multiple sets of visually complex, novel characters by comparing three learning conditions (use of air writing, paper-and-pen copying, and visual inspection only) and found a small but statistically significant advantage of air writing in accuracy of recall over the other learning conditions. While it was surprising that air writing outperformed the standard practice of writing on paper, further analysis revealed that the advantage of air writing resided in sets of characters that the students perceived difficult to memorize. This finding was consistent with the observation that the more difficulty L2 students encountered in recalling kanji, the more they used air writing (Thomas, Reference Thomas2013). Likely, the effects of air writing, which is a highly flexible and adaptable memorization technique that can be adopted in settings where other conventional techniques are unavailable, emerge when overall difficulty in a memorization task approaches a threshold.

Pedagogically speaking, an important question is whether or not students are able to improve their abilities to use available resources and strategies for learning as they advance in L2 learning, since vocabulary research consistently shows that use of multiple sources of information leads to better understanding of novel words than overreliance on a single strategy. Ulambayar (Reference Ulambayar2013), for instance, examined the effectiveness of nine-month metacognitive instruction that encouraged Mongolian novice learners of Japanese as a foreign language (JFL) to use multiple information sources (e.g., semantic radicals and contextual information) while interpreting novel kanji words. The students who received the strategy instruction demonstrated increased accuracy in their inferences, a greater number and wider range of information sources used while inferring, and fewer unknown characters that were ignored, compared with peer students who did not receive the instruction. Even without explicit metacognitive instruction, Kubota (Reference Kubota2017) reported that English-speaking undergraduate students continued to increase their kanji knowledge over five months after returning from a one-year study abroad program in Japan. Kubota attributed the continued improvement to their language experiences in Japan where they received positive feedback and acquired strategies suitable to their learning style and needs. Pizziconi (Reference Pizziconi2017), however, cautioned that L2 Japanese learners’ vocabulary growth was much larger during study abroad than one year after return, and that students going to Japan at a lower proficiency level appeared to benefit more than higher proficiency students in terms of absolute gains (i.e., gains without the consideration of advanced learners’ reduced ability to demonstrate improvement). While these studies have offered encouraging evidence supporting L2 Japanese learners’ ability to improve their metacognitive knowledge, specific factors affecting students’ gains (e.g., instructional conditions such as explicit versus. implicit learning, the types of experiences that maximize learning, learner characteristics including language abilities, etc.) and the effect of metacognitive growth on the subsequent learning remain uninvestigated. Further studies need to look into these areas.

1.2 Syntactic development

Investigation into L2 Japanese morphosyntactic development can be approached from two different theoretical angles, and post-2010 studies are no exception. One line of research is motivated by predictions based on linguistic typology concerning the learnability of particular types of syntactic structures. The last decade has seen a growing number of studies, mostly ones conducted in Japan or other Asian countries, which are motivated by the observation that certain structures in Japanese persistently pose challenges to learners from particular L1 backgrounds including highly proficient L2 Japanese speakers residing in Japan. Such observations inspire L2 Japanese researchers to determine the degree of difficulty of the acquisition of specific morpho-syntactic features of Japanese in relation to learners’ L1 backgrounds, highlighting structural differences between Japanese and other languages (e.g., Takemura, Reference Takemura2011; Nakamata, Reference Nakamata2013; Cheng, Reference Cheng2015; Nagai, Reference Nagai2015; Anubhuti, Reference Anubhuti2018). Another category of studies address a range of issues discussed in SLA research including acquisition paths (e.g., Grüter, Lieberman, & Gualmini, Reference Grüter, Lieberman and Gualmini2010; Kawaguchi & Di Biase, Reference Kawaguchi and Di Biase2012; Yabuki-Soh, Reference Yabuki-Soh2012; Iwasaki & Oliver, Reference Iwasaki and Oliver2014; Matsumoto & Nakayama, Reference Matsumoto and Nakayama2014), instruction and corrective feedback (Nakaue, Reference Nakaue2012; Usami, Reference Usami2013; Sugo, Reference Sugo2014; Hagiwara, Reference Hagiwara2015), and L2 syntactic attrition as an emerging area of interest (e.g., Russell, Reference Russell and Hansen2012).

Among language-specific morphosyntactic features, L2 learners’ acquisition of the Japanese tense and aspect markers, -ta and -te iru, has gained much attention because of the complexity of their semantic computations. The aspect marker -te iru encodes both progressive and resultative interpretations in a single form, and the notion of resultative state is generally more difficult to learn than the progressive meaning (Gabriele & McClure, Reference Gabriele and McClure2011). Furthermore, speakers of Chinese, a language that grammatically encodes aspect but not tense, were observed to use the past tense marker -ta in contexts in which -te iru was obligatory with some achievement verbs and used -te iru when -ta was appropriate for stative verbs (Chien & Nakamura, Reference Chien and Nakamura2010). Correspondingly, L1 Chinese students interpreted Japanese aspectual clues differently than Japanese natives when verbs and temporal adverbs provided inconsistent information (Long, Nakaishi, Ono, & Sakai, Reference Long, Nakaishi, Ono and Sakai2012). These distributional differences suggest that L1 knowledge plays a role in the learning of tense and aspect in L2 Japanese.

Similar crosslinguistic influence has been observed in other aspects of L2 Japanese syntactic development including postpositions and case marking particles (Nakamata, Reference Nakamata2013; Nagai, Reference Nagai2015), sentence processing (Yoshimura & MacWhinney, Reference Yoshimura and MacWhinney2010; Bai & Mukoyama, Reference Bai and Mukoyama2014; Mitsugi & MacWhinney, Reference Mitsugi and MacWhinney2016), and clausal subordinations using tame ni and yoo ni (Fukuta & Inagaki, Reference Fukuta and Inagaki2013). Motivated by the competition model, Bai and Mukoyama (Reference Bai and Mukoyama2014) demonstrated that Mongolian JFL students, whose L1 was typologically the same as Japanese, processed Japanese sentences presented in hiragana (i.e., a syllabary exclusively used in Japanese) faster than Chinese counterparts whose L1 was typologically different. The study further examined Mongolian-Chinese bilingual JFL learners, a population that was underrepresented in L2 Japanese studies, and found that the bilingual JFL learners processed all types of sentences, regardless of the availability of word order and case marking cues, virtually in the same manner. This suggests that linguistic knowledge possessed by monolingual and bilingual JFL learners work differently in processing Japanese sentences. Other competition model studies (Yoshimura & MacWhinney, Reference Yoshimura and MacWhinney2010; Mitsugi & MacWhinney, Reference Mitsugi and MacWhinney2016) demonstrated that the degree or reliance on case marking cues in Japanese sentence interpretation significantly varied among L2 Japanese learners with different L1 backgrounds and L2 proficiency, while replicating the previous finding that case marking was predominantly stronger for Japanese natives than other types of cues (e.g., word order, animacy, and honorific verb agreement as culturally determined cues). In particular, L1 English learners, who predominantly relied on word order cues while interpreting English sentences, were observed to fail to make full use of Japanese case marking cues to generate expectations about syntactic outcomes (Mitsugi & MacWhinney, Reference Mitsugi and MacWhinney2016). While there is a study showing that English speakers with no previous knowledge of Japanese exhibited sensitivity to case marking after a brief instruction of Japanese (VanPatten & Smith, Reference VanPatten and Smith2015), most competition model studies agree that L1-L2 structural differences significantly impact L2 Japanese syntactic processing.

This line of thought has been extensively explored in the framework of Universal Grammar (UG) in the past decades, and the notion of UG still motivates post-2010 crosslinguistic studies. A bidirectional study of learners of Japanese and English (Grüter et al., Reference Grüter, Lieberman and Gualmini2010), for instance, demonstrated uneven success in interpretations of disjunction and negation between the two L2 groups. The Japanese sentence John-wa eigo ka doitsugo-o hanasa-nai (John-Topic English or German-Accusative speak-not), for example, negates either language (a narrower, unmarked interpretational scope), whereas the English equivalent sentence, John does not speak English or German, negates both languages (a wider, marked scope). In Grüter et al., English-speaking learners of Japanese demonstrated evidence of having successfully acquired the interpretive properties of disjunction and negation in Japanese, whereas Japanese learners of English exhibited an interpretation pattern reflecting their L1 Japanese knowledge. This asymmetry was explained in terms of the notion of subset principles: L1 English speakers who had acquired a wider (marked) scope in their L1 had full access to a narrower (unmarked) interpretational scope in Japanese, but L1 Japanese speakers who had acquired an unmarked scope in their L1 did not have full access to a marked one in English.

Another UG-motivated study (Matsumoto & Nakayama, Reference Matsumoto and Nakayama2014) examined English-speaking JFL learners’ interpretations of Japanese overt pronouns [e.g., kare (the third person male), kanojo (the third person female)] that allowed coreference readings but not bound variable readings (e.g., *Minna_i-ga kare_i-no kaisha-ni modotta, meaning ‘all returned to his company’). As predicted from English pronouns that allowed both coreference and bound variable readings, L1 English learners of Japanese did not have problems with coreference readings but had difficulty with correctly rejecting bound variable readings in Japanese. Similarly, Gabriele (Reference Gabriele2010) found that English-speaking JFL students struggled with the interpretation of bare count nouns (e.g., kaado-o kaita ‘wrote card’) and exclusively assigned a telic reading (i.e., having the endpoint) to a verb phrase containing a bare count noun [e.g., ‘kaado-o kaita’ was interpreted as ‘wrote a/the card’ rather than as ‘wrote cards’ (an atelic interpretation assuming no endpoint)]. This tendency was accounted for by the properties of count nouns in their L1 English.

Acquisition path studies have also motivated L2 Japanese morphosyntactic development research, but the definition of ‘acquisition’ has not gained researchers’ consensus yet. As a result, different observations and quantitative findings were reported depending on whether an ‘acquisition point’ referred to emergence or accuracy. Iwasaki and Oliver (Reference Iwasaki and Oliver2014), for instance, argued that accuracy-based analysis might overlook an important aspect of L2 developmental patterns. In their longitudinal study, a child L2 learner showed substantial variations in his usage of Japanese verbal morphemes in respect to the point of emergence and the accuracy of rule applications. The morpheme -te for request, for instance, emerged at an early stage of development, but accuracy did not improve at all over their 21-month observations, suggesting that the emergence criteria might provide a better picture of L2 syntactic development. At the same time, as Iwasaki and Oliver noted, accuracy rates also showed an acquisition pattern depending on measurement levels applied. Similarly, Kizu, Pizziconi, and Iwasaki (Reference Kizu, Pizziconi and Iwasaki2013) found that the emergence of certain modal markers (e.g., explanatory modality no da, and sentence-final particles ne, yo) gradually increased as L2 proficiency improved, but the patterns of usage were inaccurate even after extended periods of formal instruction (over 350 hours) and naturalistic exposure (10–12 months stay in Japan). These findings pertain to the fundamental questions like what we mean by ‘acquisition’, ‘learning’, and ‘language development’, and how we should make those notions observable and measurable.

Another important question addressed in acquisition path studies concerns the complexity of L2 development paths. Many acquisition path studies rest on the assumption that learners go through a linear, fixed progression of development to acquire a certain morphosyntactic structure, but crosslinguistic studies have revealed that L2 Japanese morphosyntactic development is not that simple. Hashimoto (Reference Hashimoto2011), for instance, observed that young children learning JSL simultaneously used two ‘fill-in-a slot’ patterns to negate Japanese verbs: One was an ‘X-nai’ pattern applied to non-past verbs (e.g., ganbaru-nai, endure-non-past-negation, ‘not do one's best’), and the other ‘X-janai’ was used for other predicates (e.g., ki-ta-janai, come-past-negation, ‘did not come’). This study's observations were different from those of L1 acquisition data showing that Japanese children heavily relied on ‘X-janai’ only for any types of predicates (Clancy, Reference Clancy and Slobin1985).

In contrast, for adult L2 learners, who already have established L1 knowledge and well-developed cognitive systems, multiple factors, including the complexity of L2 target structures, L1-L2 structural differences, L2 proficiency and learner factors, are intertwined in L2 syntactic development, which yields mixed results. Kawaguchi and Di Biase (Reference Kawaguchi and Di Biase2012), for instance, observed that English-speaking JFL college students who had received comparable Japanese language instruction considerably differed in their passive construction production: Some were able to produce passive constructions only when sufficient time was given, some were expert passive users demonstrating automatized skills, and others demonstrated developing skills. In contrast, Yabuki-Soh (Reference Yabuki-Soh2012, Reference Yabuki-Soh2013) demonstrated a crosslinguistic influence on the acquisition of Japanese relative clauses: Korean learners, whose L1 relative close structures are similar to Japanese, showed a parallel distribution pattern to Japanese natives across course levels, whereas L1 English and L1 Chinese students showed different production patterns. In addition, the three different L1 groups did not necessarily start off with easy/simple constructions, according to the notion of the noun phrase accessibility hierarchy predicting the difficulty order in accordance with the grammatical role of the head noun, but produced difficult/complex ones at an initial stage of development. Similarly, Iwasaki (Reference Iwasaki, Iwasaki, Sells and Akita2017a) demonstrated that Korean learners, whose L1 exhibited a rich inventory of mimetics used in a grammatically similar way as Japanese, produced more Japanese mimetics in a more syntactically integrated fashion to describe motion events than English speakers whose L1 mimetics were typologically different. However, her other study (Iwasaki, Reference Iwasaki, Iwasaki, Sells and Akita2017b) indicated that L1 English learners used a greater number of mimetics in more varied structural contexts in their narratives than Korean speakers. Furthermore, Yoshioka (Reference Yoshioka, Iwasaki, Sells and Akita2017) argues in her longitudinal study of a Dutch speaker whose L1 is poor in mimetics that L2 proficiency and associability play a role in learning to use mimetics. These contradicting findings suggests the complexity of L2 syntactic development paths.

As far as instructional support for L2 Japanese syntactic development is concerned, support in accordance with learners’ cognitive processing capacities might be most effective. Hagiwara (Reference Hagiwara2015), for example, demonstrated that JFL students processed aurally presented morphemic elements in multiclausal sentences significantly better with pictorial information than without visual support, as measured by accuracy in sentence repetition. The effect of pictorial support, however, was found only for newly learned elements or those in the sentence-final position. These findings suggest that multimodal/multimedia learning may be effective when learners are able to process additional information, but that multimodal support does not work favorably when learners’ cognitive capacities are overloaded. The findings of Hagiwara's study, however, must be interpreted with caution since distributional frequency of the target elements was controlled. In other words, it remained unclear whether accurate repetition reflected learning or high availability in L2 input.

In order to investigate the effectiveness of input-processing versus output-based instruction, Nakaue (Reference Nakaue2012) examined comprehension and production of Japanese adjective comparisons (e.g., A-wa B-yori adjective) among Chinese novice JSL learners who received either type of instruction. Surprisingly, the results found that input-based instruction yielded a greater gain than output-based instruction not only for comprehension but also for production even with the absence of output practice. This finding could be interpreted as suggesting that instruction geared toward input-processing might be sufficient to learn relatively simple/straightforward forms, but the applicability of the finding to the learning of complex/difficult structures needs to be further investigated. One interesting finding concerning the quantity of input is Usami (Reference Usami2013), who found that input flood or enhancement instruction was not sufficient for college JSL students to learn to use the conjunction de, but that providing explicit grammar explanations was more effective. This finding agrees with most SLA studies, underscoring the importance of focusing on target L2 structures including corrective feedback. Yoshida (Reference Yoshida2010), however, pointed out that JFL students’ corrections did not necessarily indicate their understanding of the nature of the errors, although corrective feedback, whether explicit or implicit, overall led to learner repair regardless of target forms. Similarly, Sugo (Reference Sugo2014) reported that JFL students were able to make sense of recasts on verb conjugations and auxiliary verbs, but not on particle errors. In other words, the advantage of focusing on L2 structures and/or corrective feedback depends on specific structures to be learned.

Lastly, it is worth noting that a range of inquiries and methodological approaches adopted by studies on L2 Japanese syntactic development may reflect regional, instructional, and sociocultural factors that need to be taken into consideration. The degree of difficulty of the learning of specific morphosyntactic structures of Japanese in relation to students’ L1 backgrounds, for instance, is keenly examined by researchers and language professionals in Japan, where Japanese is used as the medium of instructional and social interactions. In such an immersion setting, different behaviors for particular forms among L2 Japanese learners with different language backgrounds can be easily observed, and such observations highlight structural differences between Japanese and other languages. In a JFL environment where Japanese is often taught as one of many foreign languages in classroom settings, on the other hand, studies on the learning of L2 Japanese are more likely to be inspired by SLA theories and research that are rooted in commonly taught languages. In other words, surrounding contexts in which L2 Japanese studies are conducted potentially affect researchers’ approaches to identify areas of problems and formulate questions to pursue. At the same time, examinations of language- or context-specific features can contribute to the advancement of general SLA and applied linguistics research by adding new sets of data and insights. To this end, studies on L2 Japanese syntactic development must continue to look into Japanese-specific issues from various perspectives.

1.3 Reading and writing

As in pre-2010 studies, a majority of post-2010 studies on L2 Japanese reading and writing concern the transferability of L1-based text-processing strategies and the degree to which learners use L2-based strategies. Matured L2 readers already have working knowledge of printed materials and text-approaching strategies acquired through their L1 literacy experiences while acquiring L2-specific knowledge. However, neither their L1-based nor L2-specific knowledge and skills are static, but they are transformative and individualized. Thus, L2 text comprehension and construction can be viewed as a dynamic process in which L1- and L2-based knowledge and skills interact and cooperate with each other. Reflecting this view, the studies reviewed in this section address how L2 Japanese learners draw on their existing knowledge to handle L2 Japanese texts and how their cognitive engagement plays a role while reading or writing in Japanese.

Crosslinguistic studies consistently show that L1-based knowledge and text-processing strategies significantly impact the development of L2 lower-level processing skills (e.g., word recognition and syntactic processing) that inevitably affect L2 text comprehension. In Horiba (Reference Horiba2012), for instance, Chinese JFL learners, whose L1-based syntactic knowledge was not useful processing Japanese sentences, exhibited weaker knowledge of syntactic relations between L2 Japanese words than Korean counterparts, whose L1 syntactic structures were similar to Japanese. However, correlation between the syntactic aspect of L2 Japanese word knowledge and text comprehension was far stronger for the Chinese group than for the Korean students, who used other aspects of L2 word knowledge (e.g., definitions, synonyms, subordinate-superordinate relations, etc.) to a greater extent than syntactic associations. Horiba speculated that for L1 Chinese readers, who had a tendency to rely on their L1 character knowledge and pay less attention to syntactic structures while reading a Japanese text, those with higher knowledge of syntactic associations between Japanese words were more likely to read better than those with lower knowledge in that area. In other words, while strong correlation between vocabulary knowledge and reading comprehension has been established in reading research, the relationship between the two components in L2 must be examined in terms of the depth of L2 word knowledge that is affected by L1-based knowledge.

To make full use of available linguistic knowledge, however, the availability of cognitive resources must be taken into consideration. L2 readers with limited L2 proficiency, for instance, tend to allocate a large portion of their cognitive resources to lower-level processing (e.g., literal comprehension of sentences) and have a limited capacity left for higher-level processing (e.g., identifying inferred meanings and broader logical structures). As a result, less proficient L2 readers may find it more difficult to identify logical relations between sentences than more proficient students. Furthermore, the degree of cognitive involvement required by learning tasks may also result in varied learning outcomes. Yoshizawa (Reference Yoshizawa2010), for instance, investigated L2 word learning among Chinese JFL students who read a Japanese passage in one of two glossing conditions requiring different degrees of cognitive involvement: In one condition the participants searched for word definitions from a given vocabulary list (high involvement), and in the other the target words were glossed in a margin (low involvement). The high involvement group outperformed the low involvement group in both text comprehension and vocabulary gain, while both groups were able to incidentally learn new words while reading.

Predicting that the presence of overt linguistic devices signifying logical connections would enhance L2 text comprehension, Ishii (Reference Ishii2011) examined the effect of the presence of two types of logical connective markers, dakara as a causative connective and shikashi as an adversative connective, on comprehension and processing time among Chinese intermediate JFL learners who read two-pair sentences with and without connectives on a computer screen. The results indicated that the presence of a connective, whether causative or adversative, facilitated the processing time of the second sentence, particularly at the initial phase, but it did not facilitate the accuracy of comprehension. Tagawa (Reference Tagawa2012), however, demonstrated the facilitative effect of logical structure search on text comprehension, compared to the identification of main ideas, among L1 Chinese learners of Japanese who engaged in two types of search activities while reading an expository text. The L1 recall and comprehension tests revealed that searching for logical structures first and then main ideas second produced the best recollection of superordinate connections between idea units, while the reverse order of the searching activities (i.e., main ideas first and logical structures second) had no impact on comprehension. Kitajima (Reference Kitajima2016), however, cautioned that proficiency in parsing clausal-based constructs in oral-like written texts did not guarantee the ability to process complex noun phrases in expository texts. In other words, L2 readers must acquire a set of skills that are different from the ability to process spoken language to become proficient in parsing complex syntactic structures in written texts.

As far as L2 text construction is concerned, L2 students’ writing reflects the traits of their L1 culture, the target culture, and the context in which their L2 learning takes place, while also being viewed as a means of self-expression. While L2 writers significantly improve their writing in terms of syntactic and textual complexity as they progress in their L2 studies (Hirotani, Reference Hirotani2013), they also show considerable individual differences in the degree to which they use their L1- and L2-based knowledge of text structures. In Kondo (Reference Kondo2013), for instance, one group of Uzbek JFL college students relied on the affective appeal rather than the rational appeal in their Japanese persuasion essays, which was similar to the trait of Japanese students’ argumentation, whereas another group of Uzbek students with comparable proficiency maintained the rational appeal just as they would do in Uzbek. Rinnert, Kobayashi, and Katayama (Reference Rinnert, Kobayashi and Katayama2015), however, identified both shared and contrasting L1- and L2-based text structures in English-speaking JFL writers’ argumentation essays. The study observed that the L2 writers took an active role in assessing audience, selecting appropriate text features, and reshaping the selected features in the given writing contexts. Adopting a dynamic view of transfer as the restructuring of available knowledge of languages, Rinnert et al., interpreted the L2 writers’ behaviors in terms of learner agency in deciding what previous writing knowledge to use when constructing L2 text and their decisions being influenced by individual learning trajectories. If it is the case that L2 writers pay less attention to corrective feedback on certain aspects of their L2 writing (e.g., content) than on other features (e.g., syntactic structures), as observed in Takahashi (Reference Takahashi2013), L2 writers’ decisions and judgments based on their knowledge of languages and writing experiences may play a central role in L2 text construction.

1.4 Summary

As we have seen, post-2010 L2 Japanese studies on lexical, syntactic, and literacy development from a cognitive perspective have added a substantial amount of insights into the learning of Japanese-specific features based on the analysis of structural differences between Japanese and other languages. The majority of the studies, however, remain focused on college-level English-, Chinese-, or Korean-speaking JFL/JSL students, resting on the unilingual assumption that L2 learners are far more dominant in their L1 than in L2. While the notions ‘L1’ and ‘L2’ must be revisited for further discussion, crosslinguistic studies reviewed in this section have confirmed that learners’ L1- and L2-based knowledge plays a significant role in L2 Japanese development.

At the same time, the studies have also shown that L2 Japanese learners with a comparable language background and L2 proficiency demonstrate considerable individual differences in their use of L1 and L2 knowledge and thereby in their achievement. One way to approach such individual differences is to look into specific circumstances under which different types of linguistic knowledge work cooperatively or competitively. Limited L2 syntactic processing, for instance, may or may not be compensated for by stronger word knowledge while reading an L2 test. An alternative approach is to examine non-linguistic factors accounting for individuals’ different learning behaviors, such as motivation as discussed in Section 3. Issues in individual differences become pressing agenda as L2 Japanese studies examine more diverse learners. Scrutinizing individual differences in both linguistic and non-linguistic terms may be a viable direction for L2 Japanese research.

Lastly, more rigorous discussions on the application of insights obtained from cognitive studies to actual instructional practices and social interactions outside of the classroom will be fruitful. What do data showing crosslinguistic influence on the learning of Japanese-specific structures, for instance, tell us about instructional goal setting and curriculum design for multilingual classrooms? How do the findings of studies examining the cognitive processing mechanisms of individual learners fit into the dynamics of social interactions involving speakers of different languages? Such questions help us to seek a point of convergence between studies conducted within cognitive and sociolinguistic/sociocultural frameworks.

2.1 Pragmatic development

Since the 2011 review, research on interlanguage pragmatics and comparative pragmatics has remained prolific and maintained agendas similar to those identified earlier. Post-2010 studies continue to investigate L2 learners’ use of interactional particles (Masuda, Reference Masuda2011; Funato, Reference Funato2012), honorifics and different speech styles (Iwasaki, Reference Iwasaki2010, Reference Iwasaki2011a; Ikeda, Reference Ikeda, Arita, Butler, Hauser, Horie, Mazuka, Shirai and Tsubakita2011a; McMeekin, Reference McMeekin2014; Moody, Reference Moody2014a; Taguchi, Reference Taguchi2014, Reference Taguchi2015), and hearsay evidentials (Narita, Reference Narita2012), or their execution of particular speech acts (Beckwith & Dewaele, Reference Beckwith, Dewaele, Ruiz de Zarobe and Ruiz de Zarobe2012; Wang, Reference Wang2013), in comparison to idealized L1 speakers’ models. The expanded coverage of publications in Japanese in the current review article confirmed that this line of research is not necessarily limited to studies that examine English speakers. Similar studies have been conducted in regard to non-English speakers studying Japanese (e.g., Chinese speakers in Taiwan, Funato, Reference Funato2012).

The scholars working in this line of research tend to assume that the core of pragmatic development lies in learners’ exposure to, or lack thereof, actual language use in social interaction. Accordingly, they explored whether naturalistic interaction is a more suitable arena than the classroom for facilitating pragmatic development. Among the various conditions, one of the most frequently studied contexts is study abroad (Iwasaki, Reference Iwasaki2010, Reference Iwasaki2011a, Reference Iwasaki2011b; Masuda, Reference Masuda2011; Beckwith & Dewaele, Reference Beckwith, Dewaele, Ruiz de Zarobe and Ruiz de Zarobe2012; Taguchi, Reference Taguchi2014). In addition to the comparative assessment of out-of-classroom experiences, another question pursued is whether or not classroom instruction—explicit teaching of pragmatic knowledge in particular—had any impact on the development (Kambara, Reference Kambara2011; Narita, Reference Narita2012; Moody, Reference Moody2014a). The inquiry of these studies generally begins with the identification of a particular object of learning, such as those listed above, as a target of intervention. A quasi-experimental research design or comparative analysis design was frequently employed, accompanied by some kinds of assessment tool incorporated as a way to measure progress in learning. Various methods including discourse completion task (DCT) (Beckwith & Dewaele, Reference Beckwith, Dewaele, Ruiz de Zarobe and Ruiz de Zarobe2012; Moody, Reference Moody2014a), oral production test (Narita, Reference Narita2012), metapragmatic knowledge test (Narita, Reference Narita2012), discourse analysis (Nagami, Reference Nagami, Chan, Chin, Nagami and Suthiwan2011; Taguchi, Reference Taguchi2014), and introspective interviews (Nagami, Reference Nagami, Chan, Chin, Nagami and Suthiwan2011) were also utilized to tap into different aspects of pragmatic development. DCT and oral production test, for example, are often adopted to confirm the effectiveness of instructional treatments and interventions, whereas discourse analysis is used to delineate L2 learner practices, as compared with those of L1 speakers.

Regardless, this line of research on interlanguage pragmatics, overtly or covertly, endorses monolingual standards where the goal of pragmatic development tends to be described (and prescribed) vis-à-vis the native speaker yardstick. Concerns regarding this widely accepted practice of setting monolingual native speaker models as an ultimate goal were raised by Mori and Mori (Reference Mori and Mori2011), but this tendency has remained prevalent. Ajioka's (Reference Ajioka2011) study specifically asked the question of what constitutes ‘native-likeness’ by analyzing how native speakers of Japanese residing in Tokyo, who are not academic researchers, perceived L2 Japanese speech during and after listening to audio-recorded interviews involving four learners. Ajioka found that smooth interaction and mutual understanding were the main elements responsible for the native speaker judgment of native-likeness, which contrasts with the previous assumption that grammar and pronunciation accuracy were the most significant. Although this finding may serve as an important wake-up call for practitioners who tend to be preoccupied with the development of accuracy in their classrooms, the very philosophy of having a native speaker standard as the instruction goal still persists.

2.2 Multilingual/translingual practices in social interaction

Meanwhile, as the field of applied linguistics undergoes the multilingual turn (e.g., May, Reference May2013; Ortega, Reference Ortega2013) and most recently the translingual turn (e.g., Canagarajah, Reference Canagarajah2013; Pennycook & Otsuji, Reference Pennycook and Otsuji2015; Hawkins & Mori, Reference Hawkins and Mori2018), the volume of L2 Japanese studies that question the validity of idealized native speaker models has increased considerably in recent years. These studies are less concerned with linguistic differences between Japanese and learners’ L1 and are critical toward the learners-as-deficient-speaker ideology (Firth & Wagner, Reference Firth and Wagner1997, Reference Firth and Wagner2007). Their central focus lies in the investigation of multilingual speakers’ multifaceted and transformative communicative repertoires and their unique contributions and sensible accomplishments in social interaction. To this end, a growing body of studies have examined how L2 Japanese speakers participated in communities of practice outside of the classroom, or how they were, or struggled to be, socialized or accommodated into Japanese society. These studies illuminate various aspects of multilingual and intercultural interactions that have not been brought to light as pedagogical concerns until recent years, including code-switching (Hobbs, Matsuo, & Payne, Reference Hobbs, Matsuo and Payne2010; Kurata, Reference Kurata2010; Greer, Reference Greer2013; Ikeda & Bysouth, Reference Ikeda, Bysouth, Haberland, Lønsmann and Preisler2013; Hasegawa, Reference Hasegawa2019), identity work (Meadows, Reference Meadows2010; Pasfield-Neofitou, Reference Pasfield-Neofitou2011; Zimmerman, Reference Zimmerman, Arita, Butler, Hauser, Horie, Mazuka, Shirai and Tsubakita2011; Kurata, Reference Kurata2014; Brown & Cheek, Reference Brown and Cheek2017), rapport making (Ji, Reference Ji2014; Hasegawa, Reference Hasegawa2019), emotion display (Burch & Kasper, Reference Burch, Kasper, Prior and Kasper2016), topic management (Ji, Reference Ji2014; Mori & Matsunaga, Reference Mori, Matsunaga, Greer, Ishida and Tateyama2017), and engagement in situated learning (Mori & Yanagimachi, Reference Mori, Yanagimachi, Koike and Blyth2015; Kasper & Burch, Reference Kasper, Burch, van Compernolle and McGregor2016). The past decade has also seen an increase in the range of contexts examined, including residential communities for JFL learners arranged as part of institutions of higher education (Martinsen, Baker, Brown, & Johnson, Reference Martinsen, Baker, Brown and Johnson2011; Mori & Matsunaga, Reference Mori, Matsunaga, Greer, Ishida and Tateyama2017) and various workplace types (Ebisuya, Reference Ebisuya2012; Ueno, Reference Ueno2013; Moody, Reference Moody2014b, Reference Moody2017; Mori & Yanagimachi, Reference Mori, Yanagimachi, Koike and Blyth2015; Pennycook & Otsuji, Reference Pennycook and Otsuji2015; Menard-Warwick & Leung, Reference Menard-Warwick and Leung2017), which are all productive areas of research in order to advance our pedagogical thinking and practice.

One of the most salient recent developments is the increase in studies that examine multilingual speakers’ code choice (alternation, mixing, switch, shifts, etc.) and their performative identities and orientations to contextual factors indexed by the choice. Greer (Reference Greer2013), for instance, examined a focus group discussion by multilingual teenagers living in Japan and discussed how their code management practices indexed different identities in the unfolding turns at talk. In particular, the choice of code in a word search was related to whom the talk was directed to in the multiparty interaction. Hence, the author concluded that word search was one interactional locus in which their multilingual identities became relevant. Similarly, Ikeda and Bysouth (Reference Ikeda, Bysouth, Haberland, Lønsmann and Preisler2013) examined language choice by international exchange students from various parts of the world at a Japanese university and described closely how they used both English and Japanese as a lingua franca. Ikeda and Bysouth reported that while English tended to be selected as a ‘troubleless’ communication tool for the purpose of mutual intelligibility, Japanese, the local language, was also used routinely as a lingua franca for rich creative styling, which helped create a sense of community among these students. From a slightly different viewpoint, Otsuji and Pennycook (Reference Otsuji, Pennycook, Choi and Ollerhead2018) described how university students in Sydney and Tokyo assembled multilingual and multimodal resources in their interactions within and beyond the classroom. The authors argued that their use of various semiotic resources in and out of class worked toward the enrichment of their educational experiences. These studies documented interesting realities of language use observed in globalizing universities and made a case for the importance of translingual pedagogies.

Beyond the academic setting, the negotiation of linguistic, ethnic, and racial identities has become an imminent issue in various facets of daily communication as the presence of foreign nationals and multilingual speakers has become prominent in Japan. Despite the fact that Japan has increasingly become multiethnic and multilingual, the established ideology of a monolingual monoethnic nation still remains dominant. Tsurutani's study (Reference Tsurutani2012), for instance, reported Japanese native speakers’ tendency to evaluate Japanese spoken by native speakers more favorably than foreign accented Japanese spoken by native speakers of Australian and British English. This circumstance makes a ‘gaijin’ (foreigner) identity more or less prominent for people who are not regarded as ‘nihonjin’ (ethnic Japanese). For example, Moody (Reference Moody2014b, Reference Moody2017) examined this gaijin identity in the workplace. Based on his focal participant, who is an English-speaking Caucasian, Moody (Reference Moody2014b) reported that the gaijin identity, along with the use of English, can be exploited as a way to initiate humorous interaction with co-workers. Examining the cases of American interns who tried to conform or not conform to the norm of honorific language use, Moody (Reference Moody2017) also shared insights into the complex process of being able to ‘fit in’ as foreign workers in Japan. The findings of this study problematized the pedagogical overemphasis of native-like targets, which can indeed be detrimental to socialization and the development of a sense of belonging to the local community in some cases. Similarly, introducing metapragmatic narratives provided by multilingual, multiethnic young adults hired by Japanese schools and companies, Menard-Warwick and Leung (Reference Menard-Warwick and Leung2017) discussed how the young adults adopted translingual practice to interact effectively with their colleagues and supervisors. Pennycook and Otsuji (Reference Pennycook and Otsuji2015) also described workplaces in urban cities of Sydney and Tokyo, where creative linguistic practices of border-crossing behaviors were appreciated among multilingual speakers coming from diverse backgrounds.

The nature of multilingual workplace interactions indeed varies widely depending on industries, types of work, and foreign workers’ linguistic and ethnic backgrounds. Ueno (Reference Ueno2013), for instance, identified three areas in which Indonesian care workers at Japanese healthcare facilities struggled: specialized terminology, comprehension of rephrased descriptions, and pragmatic language usage required for caregiving. Japanese staff members, for instance, often used onomatopoeic expressions to rephrase unfamiliar terminology or the conditions of patients, but many of the foreign workers with limited Japanese proficiency found them more difficult than learning new words for the medical concepts that were already familiar to them. Those who had acquired sufficient language proficiency to carry out daily conversations with staff members still had difficulty in listening attentively to, and having satisfactory conversations with, their patients during caregiving. Additionally, Goto (Reference Goto2015) reported that dialectal forms with no resemblance to standard Japanese would cause foreign care workers as much trouble as mimetic words. Similarly, Ebisuya (Reference Ebisuya2012) documented struggles faced by Indian interpreters in obtaining information that Japanese co-workers required from Indian engineers who were not familiar with Japanese workers’ communication practices. These observations suggest that a high level of language proficiency does not guarantee tacit knowledge of the principles of social interaction and work-specific discourse, which are acquired through actual participation in language-mediated interaction. Previous research on these issues tended to rely on information gathered through interviews or fieldnotes. Microanalysis of an even broader range of workplace interactions, supported by conversation analysis, interactional sociolinguistics and others, will hopefully contribute to a fuller understanding of how such moments of non-comprehension, misinterpretation, and conflicts are occasioned and handled, and what kinds of cultural, professional, and institutional knowledge is expected from contributing workplace members. Such investigation should shed light on various challenges faced by the increasingly multilingual and multicultural society while the implications obtained thereby should effectively fuel pedagogical decisions and classroom practices.

2.3 Multilingual youth in Japan and abroad

Alongside the general multilingual movement in applied linguistics research, the acute increase of multilingual children in Japan is an imminent social and educational issue. As the number of non-Japanese children requiring JSL instruction increased over the past ten years (MEXT, 2016), a growing body of JSL literature has discussed the language development of these minority children who were born in, or immigrated at an early age to, Japan (Castro-Vázquez, Reference Castro-Vázquez2011a, Reference Castro-Vázquez2011b; Tsutsumi, Reference Tsutsumi2012; Ikeda & Bysouth, Reference Ikeda, Bysouth, Haberland, Lønsmann and Preisler2013; Nishikawa, Reference Nishikawa2014; Namekawa, Reference Namekawa2015; Nishikawa, Aoki, Hosono, & Higuchi, Reference Nishikawa, Aoki, Hosono and Higuchi2015; Ikeda-Miura, Reference Ikeda-Miura2018).

Nishikawa et al. (Reference Nishikawa, Aoki, Hosono and Higuchi2015), for instance, demonstrated school-age JSL students’ limited knowledge of native Japanese verbs that were rarely used in school contexts and discussed the influence of L1 on their understanding of such words. Their follow-up study (Nishikawa, Hosono, & Aoki, Reference Nishikawa, Hosono and Aoki2016) further revealed that JSL children struggled more than monolingual Japanese children in acquiring action verbs with which they have little experience. Interestingly, a similar observation was made by Nishikawa (Reference Nishikawa2014), who studied JSL children who had rich and sustained exposure to Japanese from a young age. These early L2 starters—while ostensibly sounding like ‘native speakers’ in the author's eye—produced qualitatively different relative clauses from those produced by monolingual Japanese children.

One may conjecture that such differential knowledge of Japanese possessed by JSL children might adversely affect their academic language and conceptual development. Namekawa (Reference Namekawa2015), however, offered an example that may counter such a conjecture. Her focal student—a Chinese-speaking second-grader—was able to develop abstract concepts, such as shokunin-no iji (craftsman's pride), by engaging in interaction with her mother in Chinese with the help of a Japanese supporter and likening those concepts to her daily life experiences with her family and in school. This study highlighted the importance of creating a nurturing environment in which JSL children can exchange ideas with others in both L1 and L2 in order to connect newly learned concepts to their own life experiences.

The processes of JSL children's language development are shaped by their parents’ diverse circumstances, histories, and perceptions. Castro-Vázquez (Reference Castro-Vázquez2011a, Reference Castro-Vázquez2011b), for instance, interviewed Latin American families living in Japan with the aim of better understanding their life situations. The interviews illuminated the idea prevailing among foreign-born parents that they want their children to be ‘educated citizens.’ Some mothers viewed their children's schooling in Japan to be the crucial means to breaking out of apparently pervasive ethnic constraints. For them, schooling and academic certificates would grant not only social prestige but also access to first-class education, housing, health care, and employment. Thus, learning Japanese was seen as obtaining the opportunity for socio-economic upward mobility.

Multilingual youth who speak Japanese as part of their repertoires also include Japanese as a heritage learners (JHL) who reside outside of Japan. However, their experiences, of course, are not identical to JSL children given the difference in the status of Japanese in their respective environments. Namely, for JSL children, Japanese is the dominant societal language that they need to acquire to live, study, and work in Japan, whereas for JHL children, Japanese is their home language that they often struggle to develop or maintain. Different labels such as home-background speakers (Koshiba & Kurata, Reference Koshiba and Kurata2012), mixed-heritage background speakers (Shin, Reference Shin2010), linguistic minority children (Oriyama, Reference Oriyama2011), culturally and linguistically diverse children (Nakajima & Sano, Reference Nakajima and Sano2016), and transnational children (Langager, Reference Langager2010; Shao-Kobayashi, Reference Shao-Kobayashi2013) are used to refer to JHL learners, who exhibit substantial individual differences in their linguistic and ethnocultural repertoires. Regardless of the labels, however, the majority of them identify themselves as neither fully Japanese nor native citizens of their residence country and demonstrate varying levels of heritage speakers’ traits (Mori & Calder, Reference Mori and Calder2015). As a result, JHL students often do not fit into the conventional JFL paradigm that is mainly designed for students who do not have much background in Japanese language and culture (Oguro & Moloney, Reference Oguro and Moloney2010, Reference Oguro and Moloney2012).

To explore JHL learners’ language development, various factors, including age, length of residence, age of arrival, language abilities, and familial, educational, and sociocultural variables, have been investigated (e.g., Oriyama, Reference Oriyama2010, Reference Oriyama2011; Doerr & Lee, Reference Doerr and Lee2012; Hayashi & Murphy, Reference Hayashi and Murphy2013; Mori & Calder, Reference Mori and Calder2013, Reference Mori and Calder2017; Kurata, Reference Kurata2015; Nakajima & Sano, Reference Nakajima and Sano2016; Liu, Reference Liu2018a). Mori and Calder (Reference Mori and Calder2013), for example, demonstrated that JHL high school students who came to the United States by age nine were three grades behind in Japanese vocabulary knowledge and were at their grade level in English, whereas those who arrived at age ten or older were at their grade level in Japanese and two to five years behind in English. Interestingly, the negative correlation between the two languages became statistically nonsignificant when arrival age was factored out. This suggests that arrival age was a dominant factor, and that language knowledge developed by age nine seemed to be crucial in accounting for JHL learners’ bilingual development. However, Hayashi and Murphy (Reference Hayashi and Murphy2013), who examined morphological awareness of Japanese-English bilingual children in the United Kingdom, reported a somewhat contrastive finding: High knowledge of inflections and derivations in Japanese was associated with high morphological awareness in English, and vice versa. Interestingly, such bidirectional transfer was observed only for children who started to learn Japanese as L1 in Japan prior to their arrival in the United Kingdom, but not for JHL children who grew up in a bilingual family where only the mother spoke Japanese. Based on the results, the authors suggested that bidirectional transfer only occurs when the development of Japanese as L1 reaches above a certain proficiency threshold.

This line of research supports the notion that building a solid foundation in one language is important for young children to acquire literacy in multiple languages. Hence, one of the goals of JHL research involving multilingual youth is to identify familial, educational, and sociocultural factors contributing to the creation of environments where multilingual literacy is nurtured (e.g., Lauwereyns, Reference Lauwereyns2011). Mori and Calder (Reference Mori and Calder2017), for instance, demonstrated that parental support and their effort to establish good family reading habits had a positive impact on their children's development of Japanese as a home/heritage language, and that supporting home language literacy did not in the least hinder the development of the mainstream language (i.e., English in this study). Further, they reported that JHL development was influenced by parental involvement, whereas English proficiency was accounted for by families’ future residency plans and the amount of interaction by the children in English. Beyond familial factors, Oriyama (Reference Oriyama2010, Reference Oriyama2011) found that community ethnic schools fostered the formation of positive multiethnic identity as Japanese-Australian and JHL development. Oriyama argued that formal schooling in Japanese had the potential to override negative influences from a social milieu which is largely characterized by extensive exposure to the mainstream language. Doerr and Lee (Reference Doerr and Lee2012), however, pointed out that individual students and their parents viewed the same JHL school program differently, and that such perceptual differences would lead to different school experiences and subjectivity or identity constructions. These findings point to the complexity of young learners’ multilingual development as being intertwined with individual, familial, and socio-educational factors.

2.4 Summary

As reviewed in this section, the past decade has seen a surge of research that demonstrates multilingual/cultural and translingual/cultural orientations while the acquisition of Japanese-specific pragmatic norms of idealized native speakers by L2 learners has also continued to be an important research agenda. In the meantime, the mainstream pedagogical discourse appears to have embraced the latter and to have not sufficiently accounted for the realities of social interaction constructed among diverse speakers of Japanese. Moving forward, it is predicted that the kinds of studies reviewed in this section will help reconceptualize classroom practices in the coming years. To this end, important questions remain to be explored: How can we make sense of these research findings and deal with them in actual instructional practice? What should/would/could we teach for students to become competent multilingual/translingual speakers? How can we handle vast differences in individuals’ dispositions and circumstances? Answers to these questions likely differ depending on where and to whom the instruction is to be delivered. The issues in question—especially those surrounding multilingual youth—go beyond mere instructional materials or teaching techniques. We believe that these issues should be tackled as collective efforts of various entities, including policy makers and educator communities beyond Japanese language education. As shall be shown in the following sections, however, the vast majority of studies on learner motivation and teaching practices still remain focused on the classroom in higher education, presenting a significant gap in the literature.

3.1 Learner motivation

Recent developments in the study of motivations and other affective variables include the expansion of research scope from a focus on individual characteristics to an examination of the dynamic nature of perceptual development, interactions with others, and sociocultural contexts in which L2 development takes place. This is because perceptual constructs, once treated as primarily psychological, have been reconceptualized as multifaceted, dynamic entities in which a set of internal and external variables interact over time. As a result, motivational development has been captured through an examination of the context in which L2 learning is situated (e.g., Ushioda, Reference Ushioda2013; Dörnyei, MacIntyre, & Henry, Reference Dörnyei, MacIntyre and Henry2015; Ushioda & Dörnyei, Reference Ushioda and Dörnyei2017). Reflecting the current dominant status of English as a global language, however, over 70% of the studies on motivation focus on learning English as a foreign/second language (Ushioda & Dörnyei, Reference Ushioda and Dörnyei2017). In the age of globalization where diversity and inclusion are valued, there is need for the investigation of motivations driving the learning of languages other than English used in different regions across the world. Responding to this call, a growing number of L2 Japanese studies address the dynamic nature of perceptual development in relation to individual, educational, sociocultural, and regional variables in motivation (Oda, Reference Oda2010; Onishi, Reference Onishi2010; Yang, Reference Yang2011; Northwood & Thomson, Reference Northwood and Thomson2012; Kondo-Brown, Reference Kondo-Brown2013; Rose & Harbon, Reference Rose and Harbon2013; Tanaka, Reference Tanaka2013; Mori & Takeuchi, Reference Mori and Takeuchi2016), self-systems (Kurata, Reference Kurata2015; Sakeda & Kurata, Reference Sakeda and Kurata2016), perceptions (Sakurai, Reference Sakurai2012; Yoshida, Reference Yoshida2013a, Reference Yoshida2013b), attitudes (Jo, Reference Jo2010; Zhang, Reference Zhang2013), anxiety (Chiba & Morikawa, Reference Chiba, Morikawa, Chan, Chin and Suthiwa2011), and perceived learning strategies (Gamage, Reference Gamage2011; Grainger, Reference Grainger2012, Reference Grainger2013).

Based on its 2012 worldwide survey of institutions offering Japanese language education, the Japan Foundation (2013) identified a global trend of interest in manga, anime, and J-pop, etc. serving as an incentive for learning the Japanese language. The survey results indicated that overall gaining knowledge about popular as well as traditional Japanese culture were more prevalent than utility-based motivations such as future employment and study in Japan. While many L2 Japanese studies have also referred to the significance of popular culture as a common initial motivator for the study of Japanese, they have illustrated how complex and dynamic factors are at play for the retention of the interest.

Northwood and Thomson (Reference Northwood and Thomson2012), for instance, found that Australian students who continued studying Japanese exhibited a stronger integrative motivation and interests in Japanese culture, popular culture products (e.g., anime, manga), and a trip to Japan than those who discontinued. Consequently, the continuers displayed positive attitudes toward L2 Japanese learning and active engagement in self-initiated activities including practicing outside the classroom. The study, however, cautioned that, while the enjoyment of Japanese popular culture was at least partially responsible for retention in Japanese, some learners abandoned formal study of Japanese at a relatively early stage once their consumption of popular culture products was satisfied. At the same time, the continuation of formal classroom learning does not necessarily guarantee the students’ active engagement in learning. Kondo-Brown (Reference Kondo-Brown2013), for instance, observed that students in upper-level courses at an American university invested less effort in their coursework, despite their higher appreciation of socio-economic and integrative values in learning Japanese, than those at the beginning of the program. Similarly, Xu (Reference Xu2018) found that some Japanese major students in China became demotivated in the course of their study. The notable demotivating factors included the inability to achieve a desired level of proficiency, heavy course load, loss of confidence and interest, and future anxiety. These findings may suggest the importance of adapting the language curriculum to accommodate the needs and interests of students who aspire to prepare themselves for diverse personal, local, and global contexts.

In fact, motivation studies suggest that the structure and the curriculum of an academic program impact the types of students who are drawn into the program. In Taiwan, for instance, Yang (Reference Yang2011) reported substantial differences between college students majoring in Japanese and non-majors in their motivations and strategy use. Specifically, Japanese majors exhibited a higher degree of Japan-related motivation, interest in Japanese culture, and interest in popular culture than non-majors, and these motivational orientations were associated with a greater use of multiple strategies including asking for help, the acquisition of a sense of self-efficacy (i.e., belief in one's own abilities), and achievable goal setting. Furthermore, student motivation and attitudes toward Japanese language and culture continue to change over time even after enrolling in a specific program. Zhang (Reference Zhang2013) reported that Japanese majors in China increasingly exhibited positive attitudes toward Japanese people and culture with advancement to higher classes, and that such attributes were associated with increased flexibility in communication with people of different cultures. In Ukraine, where Japanese was a popular college major discipline, Onishi (Reference Onishi2010) found that the motivations of lower-year and upper-year JFL students were structured differently. Notably, first- and second-year students’ strong desire to contribute to one's own country's development, which was not identified from the upper-year students, was compelling, whereas the upper-level students showed stronger usage-based motivations than the lower-year students. In addition, career-oriented and culture-based motivations were positively correlated for both groups of students, indicating that individual students could have multiple reasons, both instrumental and integrative orientations.

To appreciate learner motivations and perceptions documented in these studies conducted in different countries, sociocultural and geopolitical circumstances of the respective world regions must be taken into consideration. What is more, the increased global mobility of students and resulting diversity in the classroom adds further complexity to this inquiry. In Australia, for instance, Grainger (Reference Grainger2012, Reference Grainger2013) found that multilingual Asian students used more strategies in different ways than Australian JFL students, suggesting the impacts of cultural backgrounds, L2 learning environment, and individuals’ multilingual capabilities on the choice of L2 Japanese learning strategies. Similarly, Mori and Takeuchi (Reference Mori and Takeuchi2016) investigated the impact of a notably increasing number of international students from China upon Japanese language courses at an American university. Their interviews with students in third- and fourth-year Japanese courses revealed that while American and Chinese students share some common reasons and goals for studying Japanese (e.g., interest in popular culture, desire to consider a post-graduate career in Japan), there are some notable differences in how they relate to Japanese language and culture, reflecting the three countries’ historical, political and sociocultural relationships. The study also addressed how the presence of international students creates both benefits and challenges. While the increased cultural diversity could foster meaningful intercultural interactions, vast knowledge of kanji and written vocabulary possessed by Chinese students, unfortunately, were viewed by some of the American students as detriments to their own learning.

Indeed, the study of kanji by native speakers of a non-character language is often seen as a challenge, which could contribute to some advanced students’ struggle to stay motivated, control emotions, and manage commitments. Rose and Harbon (Reference Rose and Harbon2013), for instance, observed that advanced L2 Japanese learners with alphabetic language backgrounds were more prone to a loss of self-regulation, defined as the extent to which learners are proactive in controlling their own learning experiences (Dörnyei, Reference Dörnyei2005), as they learn more complex characters. The study of kanji is not necessarily trouble free for L1 Chinese learners of Japanese, either, and their primary challenge seems to be to learn the pronunciation of kanji. Tanaka (Reference Tanaka2013) found that intrinsic motivation (i.e., sense of pleasure in learning) and the perceived importance of Japan predicted higher motivation and kanji proficiency for Chinese students. Conversely, introjected regulation (i.e., motivation regulated by pressure or an externally imposed value) and negative attitudes toward L2 speakers could lead them to amotivation (i.e., a lack of motivation) and lower kanji proficiency. Given that many L2 Japanese students possess the ability to handle kanji learning tasks in a systematic way, as Gamage (Reference Gamage2011) observed, what is important to acquire high kanji knowledge may not be one's ability itself but one's positive perceptions about the target language that serve as a driving force in sustaining motivation and self-regulation.

Language learning beliefs and a sense of self-efficacy also change as students accumulate learning experiences in the classroom. In Yoshida (Reference Yoshida2013a, Reference Yoshida2013b), who examined the perceptual development of four JFL college students in Australia, three learners who initially avoided speaking Japanese in class due to a lack of confidence gradually gained better self-concepts by making an effort to speak more Japanese with their classmates and manage embarrassment caused by making a mistake. The other student, who was already confident in her own speaking skills, in contrast, considered interactions with her classmates unhelpful in improving her speaking abilities and thereby did not increase her positive self-concepts. Likewise, Jo (Reference Jo2010) demonstrated that English-speaking students with negative attitudes toward a given speaking task (shadowing activities) failed to develop self-confidence in their own speaking abilities, regardless of their L2 proficiency. In contrast, a student who evaluated highly the effectiveness of the task improved his/her own abilities to shadow. Such associations between learner affect and success in learning could be interpreted as a reciprocal process: Positive attitudes toward L2 learning increase the chance for success in learning, and successful experiences promote a sense of achievement and positive self-concepts. Then, an important question to pursue is how to create such a virtuous cycle between promotion of self-confidence and successful learning in L2 instruction (Chiba & Morikawa, Reference Chiba, Morikawa, Chan, Chin and Suthiwa2011).

Teacher beliefs have also been an object of study because teacher perceptions are believed to impact their instructional decisions and assessment of individual students’ abilities (Yoshida, Reference Yoshida2010) that, in turn, significantly affect students’ expectations for their language teachers (Sakurai, Reference Sakurai2012). Yamada (Reference Yamada2014), who investigated the perceptual development of two Japanese language instructors over three years, found that beliefs that were formulated at the initial stage of their profession, and had a lot in common with other beliefs, remained strong and consistent, whereas beliefs that were formed within a short period of time and differed from other beliefs changed or faded away. While most beliefs studies have agreed that individuals’ perceptions are malleable, not all aspects of beliefs are equally modifiable: Some beliefs are more difficult to change than others.

Finally, as one of the few studies that shed light on an atypical learner, Oda (Reference Oda2010) examined how an English-speaking autistic college student was motivated to learn L2 Japanese. Learners with autism are often seen as having greater difficulty in socialization and language acquisition, resulting in insufficient opportunities to learn a foreign language. The student in the study, however, gradually showed her interest in learning Japanese in response to the tutor's approaches that took into consideration autistic students’ behaviors and perceptions and the learners’ personal interests. The study's adaptation so as to accommodate to the needs of individual students and create a nurturing environment in which the learner comfortably engaged in learning tasks offered useful implications for language education not only for those with special needs but also for L2 learners in general.

3.2 Summary

As we have seen, the majority of post-2010 L2 Japanese motivation studies emphasize the dynamic nature of perceptual change in relation to various individual, educational, and sociocultural factors including individual needs and interests, ethnolinguistic backgrounds, learning goals, L2 learning experiences, instructional and curriculum settings, and regional variables. Methodologically speaking, however, many of the studies, except for a few case studies, take a cross-sectional, quantitative approach rather than tracing a limited number of individuals’ perceptual change over a period of time. Consequently, in-depth examinations of individual learners, their personal growth or demotivation, and their surrounding contexts tend to be missing, and information about how motivational change relates to the subsequent learning remains insufficient. A single study may capture only a snapshot of motivational development, but a collection of conceptually and methodologically robust studies examining diverse learner populations and contexts should be able to untangle the complexity of L2 Japanese motivation.

4.1 Corpus-based studies and the development of materials and resources

The development of instructional materials and resources that introduce the basics of the language and models of its use is one of the fundamental steps in language teaching. Compared to what is available for the study of English by speakers of a wide variety of languages, the quantity and quality of materials and resources for learning Japanese are still in the formative stages, especially for non-English speakers. One of the notable recent trends in applied linguistics research in Japan is its emphasis on corpus-based studies and instructional materials development (Imai, Reference Imai2012; Sangawa & Erjavec, Reference Sangawa and Erjavec2012; Sunakawa, Lee, & Takahara, Reference Sunakawa, Lee and Takahara2012; Sakoda & Hosoi, Reference Sakoda and Hosoi2018). Many of the large projects initiated in Japan were generally funded by Japanese governmental agencies, through the Grants-in-Aid for Scientific Research (Kakenhi), or as part of the National Institute for Japanese Language and Linguistics initiative. The corpus-based research has various outlets in terms of its practical applications. Linguistic/discourse analysis based on corpora can be put into use for obtaining a better understanding of real-world language usage by L1 speakers (e.g., Mogi, Reference Mogi2012; Pardeshi et al., Reference Pardeshi, Imai, Kiryu, Lee, Akasegawa and Imamura2012) and of L2 speakers’ use of collocations (e.g., Terajima, Reference Terajima2016; Liu, Reference Liu2018b; Sakoda & Hosoi, Reference Sakoda and Hosoi2018). Further, corpora may serve as direct resources in developing learning materials, such as learner dictionaries and other instructional/learning tools.

Sunakawa et al. (Reference Sunakawa, Lee and Takahara2012), for instance, discussed their project of constructing a generic online database that can be used for compiling bilingual learner dictionaries in different languages. As the authors problematized, although the number of learners studying Japanese has increased across the world over the past decades, Japanese language education as a field is by no means a large market, which makes it challenging to embark on these types of initiatives that require tremendous human and financial resources. As an attempt to respond to this dilemma, the authors formed a team of 77 researchers across the world and took on the project of constructing a database that consists of various types of information, including collocations, synonyms, cultural and linguistic information in respective languages, and L2 learner corpora. The database is meant to be shared and used freely by other dictionary editors working with various languages. Similarly, Imai (Reference Imai2012) presented his project of constructing a learner dictionary of polysemous Japanese words. As the author noted, learner dictionaries are ‘conspicuously absent’ (p. 63), compared to the abundance of such resources available for learners of English. Compilation of Japanese learner dictionaries in other languages, including Italian (Mariotti & Mantelli, Reference Mariotti and Mantelli2012) and Slovene (Sangawa & Erjavec, Reference Sangawa and Erjavec2012), was also reported, indicating the shortage of such resources for learners of Japanese across the world.

Apart from learner dictionaries and resource books, corpora were used as a resource database for developing learning software. For example, Hodoscek and Nishina's (Reference Hodoscek and Nishina2012) comprehensive system development project included a writing support component, named Nutmeg, which provided learners with error correction and feedback, as well as linguistic analyses of the learners’ writings. The error correction function was made possible by integrating the learner language corpora into the system's database. There is much room for applications in this area, including the development of textbooks (e.g., Yabuki-Soh, Reference Yabuki-Soh2013). While the vast majority of existing textbooks tend to rely on example sentences and dialogs created based on material developers’ own introspection of language use, a growing number of linguists who take usage-based approaches have advocated for the importance of representing the reality of genre- or activity-specific language use (e.g., Kabata & Ono, Reference Kabata and Ono2014; Iwasaki, Reference Iwasaki2015; Matsumoto, Reference Matsumoto2016; Hudson, Matsumoto, & Mori, Reference Hudson, Matsumoto and Mori2018; Masuda, Reference Masuda2018) or concepts (Ohta, Reference Ohta2017). The integration of a fine-grained qualitative linguistic analysis and a systematic analysis of a large database enabled by technology is considered essential for quality instructional material development.

These projects are all much needed additions to the field, which offer potential for responding to the diversified learner populations and their needs. The development of genre-specific dictionaries and materials is a good example (e.g., Hobbs et al., Reference Hobbs, Matsuo and Payne2010; Oguro & Moloney, Reference Oguro and Moloney2012; Ikeda & Bysouth, Reference Ikeda, Bysouth, Haberland, Lønsmann and Preisler2013; Rose & Harbon, Reference Rose and Harbon2013). While we have observed an emerging body of research that investigated aspects of L2 Japanese acquisition using corpus technology (e.g., Ohga, Reference Ohga2017; Liu, Reference Liu2018b; Masuda, Reference Masuda2018; Sakoda & Hosoi, Reference Sakoda and Hosoi2018; Noda & Sakota, Reference Noda and Sakota2019), we hope to see more of these types of studies in the coming years. Such undertakings would likely generate a theoretical discussion that contributes to a broader field of applied linguistics and language education at large.

4.2 Instructional designs and classroom learning process

Traditionally, the classroom has been considered an essential venue where teaching and learning take place. It is where teachers present their pedagogical rationales and beliefs by setting learning goals, identifying materials, designing activities, and providing feedback on students’ performance. At the same time, students also demonstrate their own agency and preference as learners by responding in one way or another to the learning conditions provided by teachers. Studies to be reviewed in this subsection, thus, explore the convergence and divergence between teachers’ intentions and expectations, on the one hand, and students’ perceptions, experiences, and gains, on the other.

The current review yielded an observation that while some common trends in pedagogical practices can be found across world regions, types of classroom activities that are commonly adopted and hence investigated show some variations from one world region to another. For instance, a number of studies on the implementation of extensive reading, or reading for pleasure as opposed to accuracy-oriented intensive reading, has been reported by researchers in New Zealand (De Burgh-Hirabe & Feryok, Reference De Burgh-Hirabe and Feryok2012, Reference De Burgh-Hirabe and Feryok2013; Tabata-Sandom, Reference Tabata-Sandom2013, Reference Tabata-Sandom2017). Although pedagogical reports on extensive reading have been frequently delivered at national and regional conferences and seminars in North America as well, a notable trend in published studies of North American classrooms has been an increased awareness toward critical literacy education, which we will discuss further in Section 4.4. Likewise, whereas discussions on extensive reading have appeared in Japanese publications (e.g., Mikami & Harada, Reference Mikami and Harada2011), the kinds of topics enthusiastically explored in Japan may include the incorporation of can-do statements into instructional design (e.g., Morimoto, Shiozawa, Komatsu, Sekiji, & Shimada, Reference Morimoto, Shiozawa, Komatsu, Sekiji and Shimada2011; Nishigori, Reference Nishigori2019), the reconceptualization of instructional grammar according to the yasashii nihongo (easy Japanese) framework (e.g., Carroll, Reference Carroll and Gottlieb2012; Noda, Reference Noda2014; Iori, Reference Iori2015), among others.

Despite some of the possible differences noted above, by and large, emphasis on linguistic forms with a focus on proficiency elements, most notably, accuracy, fluency, and complexity, remains a general pedagogical trend in the Japanese classroom across the regions. Consequently, the majority of studies have focused on the effectiveness of particular activities and feedback methods to this end. Indeed, corrective feedback and its impact on the acquisition process was probably one of the most prominent research topics (e.g., Egi, Reference Egi2010; Yoshida, Reference Yoshida2010; Ikeda, Reference Ikeda2011b; Takahashi, Reference Takahashi2013) in the first half of the 2010s, corresponding to the general SLA trend that continued from the 1990s. A central concern of feedback research is how the intended feedback was actually perceived and understood by learners. As such, earlier feedback studies were associated with the psycholinguistic SLA theories, such as noticing and awareness, but the research scope has since expanded to include various modes, variables, and contexts. For example, while studies on corrective feedback generally tended to focus on teachers’ verbal feedback, a recent move to applying multimodal microanalytic approaches to the study of classroom interaction shed light on how teachers coordinate verbal feedback and non-verbal conduct such as posture shifts and hand gestures (e.g., Ikeda, Reference Ikeda2011b). Types of teachers’ corrective feedback have also been studied in the context of second language writing (e.g., Takahashi, Reference Takahashi2013). Further, peer feedback—in contrast to teacher feedback—has also been an object of study. Peer feedback was investigated both in the normal classroom context (e.g., Wang, Reference Wang2014) and in technology-mediated interaction, such as a Skype-based tandem project (Akiyama, Reference Akiyama2017). Hence, by and large, the central role of feedback in the learning process—how it is provided to and taken in by learners—is embedded in various aspects of Japanese language education research.

Another popular topic investigated in the past decade includes various functions of peer interaction in the classroom learning process. The use of peer or small-group work with the aim of facilitating active learner involvement is now considered standard practice. However, peer interaction can be implemented in diverse ways, and as such, close analysis of peer interaction can renew our understanding of this format of activity. For example, Hasegawa (Reference Hasegawa, Greer, Ishida and Tateyama2017) examined pair work activities in a second-semester course at a U.S. university and analyzed cases of search activity occasioned during peer interaction. His analysis demonstrated that learners were not necessarily looking for a particular lexical item—as in a ‘word search’ found in mundane conversations—but a longer chunk of talk, namely, a search for what-to-say, in their respective turns. This practice indicated a process of interaction peculiar to this pedagogical context: That is, the assigned pair work activity (question-answer practice) was highly controlled with specified linguistic outcomes while the questions were asking about some hypothetical situations (e.g., if you were a woman, what would you like to do?). Consequently, the learners were taxed to come up with ideas to produce (i.e., answer to the question) as well as attending to the focal grammar items. In another study, Hasegawa (Reference Hasegawa2018) investigated the interactional workings of laughter occasioned during similarly grammar-focused interaction. The students produced laughter at some patterned sequential positions while they followed activity cues and sequences of turns to construct their conversations. A close description of the laughter revealed intricate interplays of activity design, learners’ interpretations of the activities, unfolding learner orientations at each moment of interaction, and interpersonal relationships established between the learners.

Peer interaction may also take the form of computer-mediated collaboration in today's classroom (e.g., Jung & Suzuki, Reference Jung and Suzuki2015). Kabata and Edasawa (Reference Kabata and Edasawa2011), for instance, examined tandem learning between Canadian and Japanese students, enabled by an asynchronous crosscultural bilingual communication project. They found that the project occasioned incidental learning of various aspects of language, although its amount and quality seemed to vary depending on the students’ proficiency level. Such tandem learning projects, as well as some other types of activities that involve native speakers who do not specialize in language education, have increasingly been integrated into the curriculum to increase students’ use of Japanese outside of the classroom, to have them exposed to a Japanese-speaking community of practice, and to reduce the reliance on the teacher as sole authority. Diversifying modes of instruction enabled by the new technology, as well as changing public discourse surrounding language education, continue to challenge language educators to rethink the goals of language courses and innovative approaches to attain the goals. The following two subsections, 4.3 and 4.4, expand on these issues.

4.3 Technology-enhanced language learning and new instructional models

Since the previous review (Mori & Mori, Reference Mori and Mori2011), we have observed continued developments in the application of technologies and pedagogical innovations in and out of the classroom setting (e.g., Zimmerman & McMeekin, Reference Zimmerman and McMeekin2019). Some examples of the utilization of online communication tools (e.g., blogs, tandem learning) were already mentioned in Section 4.2. While the use of such tools was a relative novelty at the time of the previous review, it has now become a common practice. What characterizes the current state of research on technology-enhanced language learning may be a broadened scope of research programs, which are concerned not only with the pedagogical feasibility (e.g., Moore & Iida, Reference Moore and Iida2010; Kabata & Edasawa, Reference Kabata and Edasawa2011; Hirotani & Lyddon, Reference Hirotani and Lyddon2013) but also with more theoretical and ideological issues surrounding the use of technologies vis-à-vis language use and development (e.g., Doerr & Sato, Reference Doerr and Sato2011; Pasfield-Neofitou, Reference Pasfield-Neofitou2011; Walker & Haddon, Reference Walker and Haddon2011). The pedagogical feasibility research is generally responsive to new technological advances and their availability for use, but they are often reflective of, or closely tied into, the emerging needs of the current state of language programs, especially at institutions of higher education in North America and Oceania where the majority of these studies have been carried out. Most notably, in these contexts, recent economic pressures that have appeared and surged on educational productivity (see Sato, Hasegawa, Kumagai, & Kamiyoshi, Reference Sato, Hasegawa, Kumagai and Kamiyoshi2017) have apparently imposed the need for use of cost- and time-efficient language instruction technology.

One such pedagogical approach that has gained growing attention includes flipped instruction. The concept of flipped instruction is rooted in the educational value of promoting learner centeredness and learner independence, and the usability of this instructional strategy with regard to its impact on the outcomes and productivity has come under increasing scrutiny. Mori, Omori, and Sato (Reference Mori, Omori and Sato2016), for example, conducted a quasi-experimental study on the feasibility of a flip approach in teaching kanji. For the flip condition, students were instructed to study online kanji materials prior to class while the non-flip condition went through traditional in-class lectures (e.g., present-practice-produce model). The post-lesson kanji tests revealed a small but statistically significant gain from the flip approach than from the traditional approach at the initial stage of kanji learning. While the advantage of the flip condition was diminished for the intermediate level (i.e., no difference was found between the approaches), the authors recommended that a flipped model could be successfully implemented in kanji instruction. Given the presumable burden of kanji learning for many, if not most, learners of Japanese who do not have backgrounds in Chinese characters, flipped instruction may add a viable option to the growing demand for productive language instruction.

Similarly, other forms of non-traditional instruction, such as distance learning and self-instruction, by means of synchronous and/or asynchronous communication tools, are being sought as efficient language instruction delivery methods, especially in the case of less commonly taught languages (Blake, Reference Blake2013; van Deusen-Scholl, Reference van Deusen–Scholl2015). Sato, Cheng, and Jourdain (Reference Sato, Cheng and Jourdain2017), for instance, reported how a fully online, elementary Japanese course, offered as a summer course at an American university, yielded a successful outcome. At the completion of the course, which was designed in accordance with the American Council on the Teaching of Foreign Languages’ (ACTFL) World-Readiness Standards for Learning Languages (ACTFL, 2015), including various task-based activities, a simulated Oral Proficiency Interview was conducted to assess the students’ oral performance. The results indicated that the students who took the online course outperformed those who took a face-to-face version of the same course. The online students also expressed positive attitudes toward the experience in learning journals and survey responses. The authors concluded that the development of an effective, sustainable online course is possible as long as a variety of communicative tasks were incorporated to mitigate possible psychological distance and to enhance the sense of co-presence.

Aside from these reports that looked at the development and the use of new instructional tools, some studies investigated the actual learning process observed in non-traditional instructional modes. For example, Suzuki (Reference Suzuki2013) took up a unique approach and examined how off-screen behaviors of a student in an online course would reveal the micro process of learning in this environment. The author discussed a case where a focal learner, Mei, frequently engaged in verbal rehearsal of words, phrases, and sentences in parallel with the ongoing teacher-student sequence when other students were nominated to speak. This ‘overlapped speech,’ produced at a normal volume, was made possible because of the technical setup of the course; that is, only the voice of a nominated student is heard through the microphone while other students could freely talk without being heard by others. The author discussed the unique benefits of the online setup that a traditional classroom may not afford (cf., ‘vicarious response’ discussed in Ohta, Reference Ohta2001), which could sensibly be utilized in order to maximize learning outcomes.

Similarly, Umino (Reference Umino, Chan, Chin, Nagami and Suthiwan2011) examined the longitudinal process of self-instructional learning through audiovisual media by a group of adult learners in London. The students met once a week to partake in group-based self-instructional sessions for 34 weeks. At each session, the students first studied the audiovisual program on their own and then engaged in group interviews to reflect upon their learning process. This group interview process was instrumental in sharing difficulties involved in the use of the audiovisual media among the learners and overcoming them by developing better learning strategies. Although the author had to leave London by week 27, the group (of three students at that point) decided to continue meeting and carried on the self-instructional sessions for themselves. As it turned out, interestingly, the group formed a semi-autonomous community of practice (Lave & Wenger, Reference Lave and Wenger1991), wherein each learner became an indispensable member and took up a unique role. The author argued for the importance of social participatory aspects of SLA even in a self-instructional mode like the case in point.

Despite the broadened scope of inquiry into technology-enhanced language use and learning, on the whole, the volume of research in each sub-area is rather small, compared to the other areas (e.g., feedback). In particular, more work should be expected on the development and examination of alternative instructional modes, such as blended/hybrid instruction, distance education, and self-instruction, given the current and future needs of language education.

4.4 Critical approaches to teaching Japanese and content-based instruction

The ‘critical movement’ in applied linguistics, which dates back to the late 1990s (Lin, Reference Lin2014), has recently gained momentum in Japanese language education, especially in North America. While Mori and Mori (Reference Mori and Mori2011) identified critical approaches to the conceptualization of ‘culture’ and ‘language’ as an emerging trend, the application of these approaches to actual teaching practices has gained traction during the past ten years. The body of literature to be discussed below can be classified into two types: a critical approach to the analysis of culture and language (Carroll, Reference Carroll, Heinrich and Galan2011, Reference Carroll and Gottlieb2012; Suzuki, Reference Suzuki2015) and a critical approach to education in language and culture, namely, nurturing of criticality as the goal of instruction itself (e.g., Doerr & Kumagai, Reference Doerr and Kumagai2012; Kubota, Reference Kubota2012; Sato, Hasegawa, Kumagai, & Kamiyoshi, Reference Sato, Hasegawa, Kumagai, Kamiyoshi, Sato, Takami, Kamiyoshi and Kumagai2015; Sato et al. Reference Sato, Hasegawa, Kumagai and Kamiyoshi2017).

The former has contributed to advance our understanding of the ways in which the concepts of language and culture are reproduced in everyday discourse. Such reproduction is so ubiquitous that people do not even recognize it and that it effectively reinforces and reengineers stereotypes, power relationships, and inequality. By alerting readers to such discourse, this line of research attempts to challenge and defy the status quo. Suzuki (Reference Suzuki2015), for example, examined the case of banal nationalism in Japan by presenting numerous examples of media portrayals in which foreign speakers of Japanese are commodified. Similarly, Carroll (Reference Carroll, Heinrich and Galan2011) discussed examples of Japanese language as an international language from various angles. She pointed out how Japanese society embraces the notion of ‘Japan’ as the simplistic equation of Japanese state = Japanese race = Japanese language = Japanese culture. The fixed ideologies toward Japanese language and culture, which were pointed out by Mori and Mori (Reference Mori and Mori2011), are by no means a thing of the past. In fact, these ideologies may be responsible for generating the naïve acceptance of native-speakerism, as discussed above in Section 2.2. This type of research has often been linked with or discussed under the classic notion of nihonjinron (‘Theories of Japanese-ness’). Nihonjinron and its critique can provide a clear framework for research. However, the discussion should not end there and needs to be extended to encompass prospects for future changes, which seems to be largely missing. In particular, more discussions on how such a critique can be incorporated into educational and political practices would benefit practitioners.

This kind of fixed ideology is not just reconstructed by the Japanese. Meadows (Reference Meadows2010) studied how learners of Japanese working on a group task co-construct an imagined Japanese national community of practice as they interacted to prepare a booth representing Japan. By throwing in and exchanging their ideas for the booth, which they had gained through their contacts with natives, the learners discussed national authenticity (i.e., what true Japan should be). By so doing, they claimed their legitimate peripheral positioning within this community and contributed to the maintenance of nationalist ideologies. Critically reflecting on this case study, the author reminded the reader of the precarious pathway that language learners experience as cultural, social, and linguistic ‘border crossers’:

Whether authoring ‘themselves into perpetual Other-ness’ (p. 108) in the target national community or becoming part of the community and thereby destabilizing the monolingual and monocultural orientations in the community, language learners inevitably play roles in the local construction of ideologies and power dynamics. As discussed in the section on social dimensions of L2 Japanese development (Section 2), more and more L2 Japanese researchers have become mindful of such dynamics and begun to address the complexity in the classroom as well.

As a matter of fact, the past decade has seen increased implementation of critical pedagogical approaches in the classroom, most notably, at universities in North America (e.g., Sato et al., Reference Sato, Hasegawa, Kumagai, Kamiyoshi, Sato, Takami, Kamiyoshi and Kumagai2015, Reference Sato, Hasegawa, Kumagai and Kamiyoshi2017). This approach, as influenced by the Frankfurt School of critical theory, encourages practitioners to work with students in order to reflect, reexamine, challenge, and respond to all aspects of the status quo, including the ideological practice of nationalism and monolingualism. The popularity and the maturation of the critical philosophy is demonstrated by the fact that published studies cover a range of topics from conceptual and ideological discussions in research to the actual implementation of critical approaches in practice. It is probably not an accident that this development corresponds to the spread of ideas that emphasize the fostering of critical thinking across the educational curriculum, such as ‘the 21st century skills’ (Partnership for 21st Century Learning, 2015) or ‘the Essential Learning Outcomes’ (Association of American Colleges and University, 2008), in world language education and in higher education more generally.

Under the circumstance, content-based language instruction (CBI) has also gained its status as a versatile pedagogical approach, which serves as a viable ground for introducing critical perspectives in practice (Kubota, Reference Kubota2012; Sato et al. Reference Sato, Hasegawa, Kumagai and Kamiyoshi2017). CBI has been popularized in JFL programs in U.S. higher education (Takami, Kumagai, Sato, Hasegawa, & Morioka, Reference Takami, Kumagai, Sato, Hasegawa, Morioka, Sato, Takami, Kamiyoshi and Kumagai2015), but the types and the degrees of CBI and critical pedagogical approaches vary considerably. There are also various subject areas covered, including history (Chikamatsu, Reference Chikamatsu2012; Kubota, Reference Kubota2012), religion and spirituality (Goulah, Reference Goulah2011), literacy (Kumagai & Iwasaki, Reference Kumagai and Iwasaki2011; Takimoto & Hashimoto, Reference Takimoto and Hashimoto2011; Sato et al., Reference Sato, Hasegawa, Kumagai, Kamiyoshi, Sato, Takami, Kamiyoshi and Kumagai2015, Reference Sato, Hasegawa, Kumagai and Kamiyoshi2017), and popular culture (Doerr & Kumagai, Reference Doerr and Kumagai2012; Suzuki, Reference Suzuki2015).

While the critical approach to CBI, or Critical Content-Based Instruction (CCBI) (Sato et al., Reference Sato, Hasegawa, Kumagai, Kamiyoshi, Sato, Takami, Kamiyoshi and Kumagai2015, Reference Sato, Hasegawa, Kumagai and Kamiyoshi2017), appears to gain acceptance among an increasing number of practitioners, there remain practical questions when it comes to its actual implementation. What constitutes criticality at the level of practice? What kinds of learning goals can be established while keeping a balance between the fostering of criticality and the improvement of language proficiency? How can the fulfillment of the set learning goals be measured through assignments and activities? What kind of training is necessary for instructors to design a successful CBI? Indeed, the effectiveness of CBI in terms of learning outcomes has seldom been investigated, with the notable exception of Douglas (Reference Douglas2017), who reported her assessment of a CBI-based Japanese language course offered at a U.S. university. According to the study, while students’ English essays produced at the end of the course demonstrated that they uniformly gained content knowledge, the degree of improvement in the quality of Japanese writing and other aspects of language ability varied depending on the students’ level of proficiency prior to the course. While confirming CBI's potentials, Douglas recommended the creative execution of differentiated instruction to maximize possible growth of each student who requires support in different areas. As higher education continues to evolve, the role, responsibilities, and expertise of language educators in academic institutions have been, and continue to be, reshaped in the new contexts. The questions posed above cannot be answered without further research on relevant language and educational policies and ideologies, as well as the accumulation of sound case studies.

4.5 Summary

Overall, we have seen the development of research areas in accordance with what, apparently, reflects the changing surroundings of Japanese language education. These pedagogical approaches and practices are obviously welcome efforts of researchers and practitioners. However, the new developments still tended to be limited to the context of higher education institutions, and relatively limited attention has been paid to other contexts, including primary and secondary education (K-12). Further, studies on language education for multilingual youth and migrant workers appear to be still underdeveloped relative to the urgent needs discussed in Section 2.3 Mori and Mori (Reference Mori and Mori2011) pointed out that studies focusing on K-12 contexts were seldom found in the type of publication venues surveyed for the review article, but there does not seem to have been much change since then—with notable exceptions of Turner (Reference Turner2013), which discussed the feasibility of adopting the European model of Content and Language Integrated Learning in Australian primary and secondary schools, and De Burgh-Hirabe and Feryok (Reference De Burgh-Hirabe and Feryok2012, Reference De Burgh-Hirabe and Feryok2013), which studied the effectiveness of extensive reading for high school students in New Zealand. This skewed coverage should be recognized by L2 Japanese researchers, who are in positions to facilitate fruitful collaborations with practitioners teaching Japanese in various circumstances. How to address issues presented by diverse language education contexts and learner populations remains a critical agenda item. Advancement of both theory and practice in the field is heavily dependent on the continued interaction of teaching and research.