1.1 Vocabulary and kanji learning

The growing number of studies investigating L2 Japanese vocabulary learning suggests that lexical development is a major concern among L2 Japanese researchers and educators. Many of those studies specifically address L2 kanji learning. For school-age or adult L2 Japanese learners who often learn new words from printed materials, learning the different types of orthography used in written Japanese (i.e. the two types of syllabaries – hiragana and katakana – and a logography – kanji]) entails the learning of different types of words (e.g. Japanese-originated words, grammar function words, foreign loan words, onomatopoeia, and Sino-Japanese words). In particular, the acquisition of a good working knowledge of kanji is essential for fluent reading, as kanji words constitute a crucial part of Japanese written vocabulary (Y. Mori Reference Mori and Hatasa2003b).

L2 Japanese learners' vocabulary knowledge is both quantitatively and qualitatively different from native norms (Komori Reference Komori2003; Matsumoto & Horiba Reference Matsumoto and Horiba2007). Furthermore, first language (L1) orthographic experience has a significant impact on the development of L2 reading sub-skills, including phonemic awareness (Navas Reference Navas2004) and word processing strategies (see Koda Reference Koda, Koda and Zehler2008 for a review). L2 Japanese learners whose L1 is phonographic (i.e. a language that employs a sound-based orthography, such as English), for instance, exhibit more reliance on phonological, and less on visual, information when recognizing words in the L2 than those with a logographic L1 (i.e. a language that uses a morpheme-based orthography, such as Chinese). Chikamatsu (Reference Chikamatsu2006), however, demonstrated that higher-proficiency English-speaking L2 Japanese learners showed more visual reliance (i.e. L2 orthographic strategies) and less phonological reliance (i.e. L1 strategies) in L2 hiragana and katakana word recognition than their lower-proficiency counterparts, suggesting that L2 word recognition strategies can be reconstructed as students gain greater L2 proficiency. Hansen, Umeda, & McKinney (Reference Hansen, Umeda and McKinney2002) findings are in accordance with this result: they found that English-speaking missionary returnees from Japan or Korea with more advanced L2 knowledge learned more new and once-learned-but-forgotten items than those with lower L2 proficiency, suggesting that L2 proficiency also has a significant impact on L2 word learning and relearning strategies.

Many words are polysemous, with their exact meaning largely determined by context. Research on incidental vocabulary learning – learning word meanings as a by-product of comprehending overall meaning – has clearly documented L2 learners' ability to learn new words from context. Adachi (Reference Adachi2003), for example, found that providing varied exemplars that showed clear distinctions between semantically similar words promoted accurate word usage. Wei (Reference Wei2006), who compared the gains of three instructional groups (reading plus vocabulary exercise, reading only, and control), demonstrated that explicit vocabulary instruction, in conjunction with reading for meaning, enhanced intermediate L2 Japanese learners' vocabulary learning. While her study found that reading-only instruction had a significant effect, students who received vocabulary instruction after reading learned more vocabulary than the reading-only group. Similarly, Kitajima (Reference Kitajima2001) and Adachi (Reference Adachi2005) documented the advantage, for retention, of using newly learned target words in meaningful interactions over receptive learning activities. These studies support the idea that multiple exposures in different contexts and explicit operations, combined with reading for meaning, could compensate for limited contextual clues, as pointed out by Y. Mori (Reference Mori2003a; Reference Mori and Hatasa2003b).

Dictionaries and/or marginal glosses are commonly used to enhance L2 comprehension and subsequent vocabulary learning. Yoshizawa (Reference Yoshizawa2005), however, points out that their effects must be considered in terms of L2 proficiency and the purposes of reading. In her study, high- and low-proficiency Chinese learners of Japanese read a short passage for comprehension under one of three conditions: under the first, students were allowed to use dictionaries, under the second, they were provided with a text with marginal gloss in Chinese, and the third formed the control group. The results showed that while both high- and low-proficiency students gained from marginal glosses for word learning, only the high-proficiency group gained from dictionary use for reading comprehension and word learning.

The effects of mnemonics (memory-enhancing strategies) through the use of personal encoding techniques or mental images of the items to be learned are another subject for debate. Mnemonic devices include keywords, visual images, and phonological resemblance to the target items; memory aids can either be formally pre-established or idiosyncratically generated by the learner. Using pictorial mnemonic strategies, Manalo, Mizutani & Trafford (Reference Manalo, Mizutani and Trafford2004) taught hiragana, katakana, and kanji characters to 27 introductory students who were experiencing difficulties with coursework. While no statistically significant improvement was observed in pre- and post-session test scores, a significant gain was observed in test-passing rates. Furthermore, the students reported that the sessions helped in their course performance and with assessments, suggesting that mnemonic strategy instruction has a positive impact on learner perceptions, rather than on test performance as such.

Kanji learning is considered one of the most challenging aspects of learning Japanese, and some students express a desire for some sort of strategy instruction (Y. Mori Reference Mori2002; Shimizu & Green Reference Shimizu and Green2002; Mori & Shimizu Reference Mori and Shimizu2007). The challenge in kanji learning includes the difficulty of retention, multiple readings of a single character, visual similarities between characters, the polysemous nature of kanji words, the large number of characters to learn, and their visual complexity. As a result, issues in L2 kanji include kanji learning/teaching strategies (Shimizu & Green, Reference Shimizu and Green2002; Mori & Shimizu Reference Mori and Shimizu2007), memory and retrieval (Chikamatsu Reference Chikamatsu, Cook and Bassetti2005), radical/compositional awareness (Toyoda Reference Toyoda2000; Yamashita & Maru Reference Yamashita and Maru2000; Kubota & Toyoda Reference Kubota and Toyoda2001; Mori, Sato & Shimizu Reference Mori and Shimizu2007), the role of contextual and morphological information in interpreting novel kanji words (Y. Mori Reference Mori2002, Reference Komori2003; Kondo-Brown Reference Kondo-Brown2006a; Mori et al. Reference Mori, Sato and Shimizu2007), the transferability of L1 knowledge among learners with Chinese-character backgrounds (Kato Reference Kato2005; Sawabe & Yasui Reference Sawabe and Yasui2008), the role of teacher feedback in learning to write kanji characters (Kubota Reference Kubota2005), and teacher/student perceptions of kanji teaching/learning (Y. Mori Reference Mori2002; Shimizu & Green Reference Shimizu and Green2002; Gamage Reference Gamage2003; Grainger Reference Grainger2005; Mori & Shimizu Reference Mori and Shimizu2007).

Using both qualitative and quantitative methods, Shimizu & Green (Reference Shimizu and Green2002) identified six factors in Japanese language educators' attitudes toward kanji (viz. cultural tradition, the difficulty of kanji, affective orientation, aptitude, the future of kanji, and its usefulness) and two elements of self-reported kanji teaching strategies (viz. memory and context strategies). Interestingly, teachers reported that rote learning was the most frequently used strategy, and their appreciation of cultural tradition in kanji and its practicality was associated with positive attitudes toward kanji teaching and the use of a variety of instructional strategies. Drawing upon this conceptualization, Mori & Shimizu (Reference Mori and Shimizu2007) identified six dimensions of L2 Japanese students’ attitudes toward kanji, represented by the following beliefs: It is fun; It is difficult; It has cultural value; It has a future; It is useful; and Kanji learning requires special abilities, and six self-reported kanji learning strategies: Morphological analysis, Rote memorization, Context-based strategies, Association strategies, Metacognitive strategies, and Helplessness. Further analysis revealed that students considered Rote memorization the most effective and Metacognitive strategies the least effective, and that positive attitudes toward kanji were associated with the use of varied strategies. The similarities between the kanji beliefs of teacher-s and students may reflect the significant impact of educators' teaching philosophies on learner perceptions.

The studies outlined above only examined students’ own view of their learning strategies, so Mori, Sato & Shimizu (Reference Mori and Shimizu2007) explored the relationship between students’ perceptions and their actual use of strategies in handling novel kanji words. The results indicated that belief in the effectiveness of metacognitive strategies could predict success in the compositional analysis of novel kanji characters/words, suggesting that enhancing students’ metacognitive awareness may increase their chances of becoming strategic kanji learners. Similarly, Kondo-Brown (Reference Kondo-Brown2006a) demonstrated moderate yet significant correlations between affective factors (perceived reading ability, perceived difficulty of learning kanji, and their level of motivation for reading) and reading proficiency and kanji knowledge. Interestingly, kanji knowledge was more closely related to perceived reading ability than to reading proficiency, suggesting that kanji knowledge plays a significant role in promoting students’ development of their own effectiveness in reading, and vice versa.

Kanji words fall into two categories in terms of their compositional structure. One type are the single-character words consisting of sets of less complex parts known as radicals. A radical provides semantic and/or phonetic information that, at least partially, contributes to the meaning and/or the sound of the character. The other type are kanji compounds: words consisting of two or more kanji characters. L2 Japanese learners are aware of the semantic and/or phonetic function of familiar parts of kanji words and can apply their compositional knowledge to the learning of new words (Toyoda Reference Toyoda2000; Yamashita & Maru Reference Yamashita and Maru2000; Kubota & Toyoda Reference Kubota and Toyoda2001). Toyoda (Reference Toyoda2000), however, points out that less information can be inferred from word components than from other sources, and that learners are better at using semantic information than phonetic information. Similarly, L2 students are able to use information from component characters to infer the meanings of novel kanji compounds (Y. Mori Reference Mori2002, Reference Mori2003a; Kondo-Brown Reference Kondo-Brown2006a; Mori et al. Reference Mori, Sato and Shimizu2007).

Information from kanji components, however, is not always straightforwardly related to the meanings of the resulting words (Y. Mori Reference Mori2003a). Consequently, L2 students often misinterpret unfamiliar words (Y. Mori Reference Mori2002) or fail to work out their meaning (Kondo-Brown Reference Kondo-Brown2006a). Students should therefore be encouraged to use multiple sources of information in interpreting novel kanji words (Y. Mori Reference Mori2002; Reference Mori2003a). At the same time, students show considerable individual differences in the way they learn new kanji words (Kubota & Toyoda Reference Kubota and Toyoda2001; Y. Mori Reference Mori2002). Y. Mori (Reference Mori2002), for instance, demonstrated that while the majority of the participants were able to combine information from kanji components and sentential context in an additive way, nearly half of the students over-relied on a single source and did not check their inferences against other sources of information. These students may be the ones who benefit from strategy instruction.

Studies of individual differences have shown that students’ choice of strategies for a challenging learning task reflects their views on language learning, in general, and the nature of the given task, in particular. Using the Strategy Inventory for Language Learning (SILL) (Oxford Reference Oxford1990), Grainger (Reference Grainger2005) demonstrated that L2 Japanese learners, like learners of Western languages, scored relatively highly on most literacy-related items of the SILL (such as guessing from context), suggesting that orthographic divergences made little difference in perceived literacy strategy use. Gamage (Reference Gamage2003), however, found that learners of Japanese as a foreign language (JFL) from alphabetic backgrounds reported that they would use repeated writing strategies to memorize kanji words more frequently than learners from Chinese character backgrounds. Future research must clarify the factors that contribute to the formulation of learner perceptions and provide explanatory accounts for the relationship between perceived strategies and actual learning behaviors.

1.2 Interlanguage grammar development

A central concern in SLA research has been L2 learners' syntactic development. L2 Japanese research is no exception, and interlanguage development has been explored in connection with various aspects of syntactic acquisition, including verb morphology (Kawaguchi Reference Kawaguchi2000; Di Biase & Kawaguchi Reference Di Biase and Kawaguchi2002; Hashimoto Reference Hashimoto2006, Reference Hashimoto2007), passive and causative constructions (Taguchi Reference Taguchi2001; Sawetaiyaram Reference Sawetaiyaram2009), negation (Kamura Reference Kamura2001, Reference Kamura2003), argument structures (Hirakawa Reference Hirakawa2001; Inagaki Reference Inagaki2001; Sorace & Shomura Reference Sorace and Shomura2001), noun-modifying clauses (Kanno Reference Kanno2001, Reference Kanno2007; Mukouyama Reference Mukouyama2004; Ozeki Reference Ozeki2005; Ozeki & Shirai Reference Ozeki and Shirai2007; Yabuki-Soh Reference Yabuki-Soh2007; Yamashita Reference Yamashita2008; Sawasaki Reference Sawasaki2009), temporal phrases (Ozeki Reference Ozeki2003), conditional expressions (Ozeki Reference Ozeki2008), postpositional particles (Hasuike Reference Hasuike2007; Hashimoto Reference Hashimoto2009), and numeral classifiers (Hansen & Chen Reference Hansen and Chen2001).

One line of research in this area focuses on testing the applicability of universal developmental paths and language-processing strategies to L2 Japanese learning. It has been observed that young novice L2 learners' oral productions exhibit syntactic and discourse features that are similar to those of L1 children and adult JFL learners, including word-level utterances, repetition, reliance on formulaic syntactic structures, difficulty with adjectival inflections (particularly negative and past forms) compared to verb inflections, use of one-to-one form-function mapping strategies, reliance on the canonical SOV word order, and difficulty with multifunctional particles and passive constructions (Antonek, Donato & Tucker Reference Antonek, Donato and Tucker2000; Donato, Tucker, Wudthayagorn & Igarashi Reference Donato, Tucker, Wudthayagorn and Igarashi2000; Igarashi, Wudthayagorn, Donato & Tucker Reference Igarashi, Wudthayagorn, Donato and Tucker2002), and that, as their proficiency increases, L2 learners produce more syntactically complex, longer utterances (Iwashita Reference Iwashita2006). These observations lead to the hypothesis that, like L1 children, L2 Japanese learners go through a fixed developmental sequence that can be predicted by the hierarchy of markedness (the degree to which a certain form is distant from a basic, default form).

Di Biase & Kawaguchi (Reference Di Biase and Kawaguchi2002) and Kawaguchi (Reference Kawaguchi2000) tested the applicability to L2 Japanese of Pienemann's (1998) verb morphology hierarchy, which predicted four developmental stages: invariant forms (e.g. tabemasu ‘eat’ as a word), lexical-semantic morphemes (e.g. tabemasu-ka ‘[do you] eat?’; tabe-masen ‘eat [polite negative]’), phrasal morphemes (e.g. nom-a+na-i ‘do not drink’), inter-phrasal morphemes (e.g. passive, causative, and benefactive [giving-receiving] forms). Kawaguchi's production data, obtained in both longitudinal and cross-sectional studies, tend to support this developmental sequence: L2 Japanese verbal morphology seems to be acquired in a fixed order that is predictable through the identification of the hierarchy of processing prerequisites for sentence production.

Studies of L2 Japanese relative clauses (e.g. Ozeki Reference Ozeki2005; Ozeki & Shirai Reference Ozeki and Shirai2007; Sawasaki Reference Sawasaki2009), however, did not support the generalizability of the noun phrase accessibility hierarchy (NPAH) (Keenan & Comrie Reference Keenan and Comrie1977, cited by Ozeki & Shirai Reference Ozeki and Shirai2007), which predicts the order of the ease of relativization as a function of the grammatical role of the head noun phrase modified by the relative clause: subject > direct object > indirect object > oblique (e.g. watashi-ga kenka-o shita-hito ‘the person with whom I had a fight’) > genitive > object of comparison. Ozeki & Shirai (Reference Ozeki and Shirai2007), for instance, found that subject relative clauses were not necessarily easier than direct object or oblique relative clauses for L2 Japanese learners, and that the animacy of the head noun played a significant role in learners' interpretations of Japanese relative clauses. They attributed these inconsistent findings to structural differences between noun-modifying clauses in Japanese and gap-involving relative clauses in European languages.

Furthermore, longitudinal studies with L2 children in Japan (Taguchi Reference Taguchi2001; Hashimoto Reference Hashimoto2006, Reference Hashimoto2007, Reference Hashimoto2009) have shown that L2 children use different learning strategies and, as a result, go though different developmental paths from L1 children. Hashimoto (Reference Hashimoto2006), who examined an English-speaking L2 Japanese preschooler's acquisition of the past-tense marker ta, found that (a) the L2 child produced non-target like utterances that were not observed in an L1 child's productions, and (b) unlike the L1 child, the L2 preschooler first generated a ‘pivot’ schema with a slot based on a cognitively and phonologically salient unit (e.g. [ ] + chatta ‘end up -ing’) and placed any unit into the slot until she started to place verbs (e.g. *dekiru+chatta *‘end up being able to’). Likewise, Taguchi (Reference Taguchi2001) observed that a Brazilian fourth-grader acquired the Japanese passive and causative constructions as a series of unanalyzed expressions (e.g. okorareru ‘be scolded’), instead of learning to use inter-phrasal morphemes productively (e.g. okor-arer-u ‘scold [passive-nonpast]’), suggesting that the syntactic development of young JSL learners in bilingual environments might be more accurately described as a non-linear, rather than a sequential, progression.

Despite these differences, research consistently suggests that L2 learners' knowledge of form-function mapping is rule-governed (Hirakawa Reference Hirakawa2001) and developmental (Sorace & Shomura Reference Sorace and Shomura2001; Ozeki Reference Ozeki2003, Reference Ozeki2008). Using the framework of Universal Grammar, Hirakawa (Reference Hirakawa2001) demonstrated English-speaking JFL learners' sensitivity to the syntactic and semantic properties associated with Japanese unergative and unaccusative verbs that display distinct argument structures in intransitive constructions. Unergative verbs are typically associated with activities, so their argument is usually an agent generated in the subject position (e.g. John-ga arui-ta ‘John walked’). Unaccusative verbs, in contrast, often denote a change of state, and their argument is a theme or patient generated in the object position (e.g. John-ga tui-ta ‘John arrived’). While Hirakawa's study demonstrated that L2 Japanese learners generally knew the properties of these argument structures, Sorace & Shomura (Reference Sorace and Shomura2001) found that unergative verbs were acquired earlier than unaccusative verbs, suggesting that that type of knowledge was developmental. In other words, intermediate JFL learners were less sensitive to the syntactic optionality of unaccusative verbs than to the non-optionality of unergative verbs. For example, they recognize the acceptability of the two sentences with the counter (or ‘numeral classifier’) san-nin (three-persons) at a different position while maintaining the same meaning: Gakusei-ga san-nin Tokyo-ni tsuita and Gakusei-ga Tokyo-ni san-nin tsuita ‘Three students arrived at Tokyo’. On the other hand, they may not recognize that while Gakusei-ga san-nin wazato warat-ta ‘Three students intentionally laughed’ is acceptable, *Gakusei-ga wazato san-nin warat-ta is not. Higher proficiency students made better judgments about the unacceptability of unergative sentences than lower proficiency students.

In addition to L2 proficiency, L1 influence, L2 input, and instructional effects must be considered in L2 Japanese syntactic development, since these factors could alter the markedness hierarchy. Inagaki (Reference Inagaki2001), for instance, predicted that Japanese-speaking L2 English learners and English-speaking L2 Japanese learners would learn manner-of-motion verbs in the target language in different ways, based on the syntactic differences between English and Japanese. English manner-of-motion verbs with a prepositional phrase indicating a goal (goal PP) (e.g. walk to school) are grammatical and thus exist in English input (providing positive evidence), whereas their Japanese counterparts (e.g. *gakkoo-e aruku ‘walk to school’) are unacceptable and thus do not exist in Japanese input (providing negative evidence). Grammaticality judgment tests revealed that, while Japanese-speaking L2 English learners correctly accepted English manner-of-motion verbs with goal PPs, English-speaking L2 Japanese learners, even at an advanced level, incorrectly accepted the Japanese counterparts despite the absence of the argument structure in Japanese, suggesting that L2 learners would rely on their L1 knowledge when an L2 argument structure had more constraints than its L1 counterpart. Kanno's (Reference Kanno2007) report, that the properties of L1 relative clauses (e.g. head direction, word order, and relative order of filler and gap) might affect learners' interpretations of difficult L2 relative clauses, is consistent with this interpretation.

Marked forms that appear frequently in L2 learning environments, however, may be acquired earlier than the markedness scale would predict. Hansen & Chen (Reference Hansen and Chen2001), for instance, unexpectedly observed that English-speaking missionaries who had stayed in Japan for a relatively short period of time made correct use, in a picture-describing task, of satsu (the counter for books) and dai (the counter for large mechanical objects), both of which were high-frequency classifiers but were considered marked because of their inanimacy. In addition, Yabuki-Soh (Reference Yabuki-Soh2007) found that first-level JFL students who received form-based, explicit instruction on marked relative clauses (e.g. oblique relativization) understood less marked items better than those who received meaning-based or form+meaning-based instruction, demonstrating L2 Japanese learners' ability to generalize rules for other types of relative clauses. These findings suggest that teaching marked items may enhance students’ attention to detailed analysis of form and thus facilitate the development of relativization in the L2.

To summarize, the studies reviewed above consistently suggest that L2 Japanese syntactic development is not random but largely predictable from general principles; however, factors such as L1–L2 structural differences, L1 influence, L2 proficiency, L2 input, and instruction must be taken into consideration, as they may have a significant impact on the acquisition paths predicted by theories. At the same time, the notion of ‘acquisition’ calls for re-examination, as individual studies used the term differently and examined different types of knowledge (e.g. production, accurate usage, comprehension, and grammaticality judgment). L2 learners' syntactic knowledge is dynamic and subject to change over time, so rigorous longitudinal, as well as cross-sectional, studies from various perspectives are needed in this area.

1.3 Reading and writing in L2 Japanese

The acquisition of solid reading and writing skills is essential for L2 learners, particularly for school-age or adult JFL learners whose primary learning resources are printed materials. Fortunately, older L2 learners have age-appropriate experience in L1 literacy, conceptual knowledge, and cognitive resources that are not usually fully available to young L1 learners. As a result, L2 learners are expected to have a working schema for written materials and their own strategies for approaching text, acquired through their L1 experience. One major issue in this area is thus the transferability of L1 text-processing strategies into the L2 (see Nara Reference Nara, Nara and Noda2003 for a review of transferable and nontransferable reading skills). This line of thought leads to the identification of factors that facilitate or interfere with the use of L1 reading or writing strategies in L2.

Cognitive overload is one factor that may hinder the use of L1 text processing strategies in L2. Due to their limited L2 knowledge, L2 readers tend to allocate a large part of their cognitive resources to lower-level processing, including character/word recognition and sentence processing. Consequently, little capacity is left for higher-level processing, resulting in the inability to make inferences or predictions, as proficient readers do in L1. For a manageable text or task that does not take much of the reader's cognitive capacity, or for high proficiency L2 readers who have acquired efficient lower-level processing skills, higher-level processing strategies might be accessible. To address this question, Horiba (Reference Horiba2000) compared native and L2 readers’ text-processing strategies as they tackled demanding or less demanding texts and tasks. The results showed that native readers generated more backward inferences and more comments on text structure for demanding texts and tasks than for less demanding ones, but that nonnative readers did not process texts differently according to the nature of the task. In other words, nonnative readers did not strategically engage in different modes of processing according to text type and task as native readers did, and their text processing was particularly inefficient for demanding texts and tasks.

If efficient lower-level processing frees L2 readers’ cognitive capacity for higher-level processing, then Chinese students, who already have good working character knowledge, should be able to read L2 Japanese texts better than non-Chinese students. Machida (Reference Machida2001a) found, however, that Chinese and non-Chinese intermediate L2 Japanese learners showed no significant differences in their text-approaching strategies, and that the non-Chinese group even outscored the Chinese group in reading comprehension tasks, while the Chinese group outperformed in kanji comprehension tasks, both with and without context. These unexpected findings suggest that L1 character knowledge does not necessarily guarantee skilled L2 reading unless readers pay attention to other sources of information in a text. The superiority of the non-Chinese group in reading comprehension, however, must be interpreted with caution, as it might be attributable to their higher L2 proficiency, rather than to their L1 knowledge.

The benefits of extensive reading, or reading for pleasure, in improving vocabulary and reading comprehension skills have been extensively documented in L1, as well as in L2 (e.g. Krashen Reference Krashen2004). To explore the applicability of these findings to L2 Japanese, Hitosugi & Day (Reference Hitosugi and Day2004) implemented a ten-week extensive reading program in a second-semester Japanese course at the University of Hawai'i. In the program, students selected what they wanted to read from a collection of Japanese children's literature and read as many books as they could at their own pace, following the guidelines for an extensive reading program given by Day & Bamford (2002). By the end of the program, the extensive-program students had read an average of 31.6 books in ten weeks, gained more in reading comprehension, and showed stronger motivation for Japanese learning than those who did not participate in the program. Similarly, Leung (Reference Leung2002) observed positive effects of extensive reading on vocabulary learning, reading comprehension, and attitude toward reading through a case study of an adult JFL learner's self-study over 20 weeks.

Writing in L2 can be viewed as an effective means of self-expression, communication, and language learning, but L2 writers often face various challenges, such as their limited L2 proficiency, a lack of L2 writing experience, time constraints, and conceptual difficulty. Hence, issues in L2 Japanese writing include: how students define L2 writing tasks (Nakayama Reference Nakayama2002; Haneda Reference Haneda2004, Reference Haneda2005), the effects of L1 and learning backgrounds (Woodall Reference Woodall2002; Iwashita & Sekiguchi Reference Iwashita and Sekiguchi2009), L2 proficiency (Kubota Reference Kubota2001), computer use (Chikamatsu Reference Chikamatsu2003; Takahashi Reference Takahashi2009), and the role of feedback (Kubota Reference Kubota2001; Kondo-Brown Reference Kondo-Brown2002; Hirose Reference Hirose2004).

Haneda (Reference Haneda2005) views L2 writing as the interplay between L2 learners' daily interactions with others and macrosocial situations, where their L2 use takes place. In her study, two Canadian university students from different ethnographic backgrounds (a Japanese heritage learner and a Caucasian student with strong career-oriented goals) exhibited considerable differences in the way they composed a text, differences that were attributable to the learners' life histories, their relations to the target language and culture, and their identities. In her other study (Haneda Reference Haneda2004), which examined the discourse patterns in one-on-one teacher-student interaction in joint writing conferences, she found that their interaction varied according to students’ self-selected revision goals, the choice of discourse topic, and the teacher's pedagogical goals. Furthermore, the pedagogical effectiveness of a teacher-student interaction came not from the formal properties of the interaction (e.g. triadic dialogue (Initiate-Response-Evaluation/Follow up) or the types of questions, but from the way in which the teacher used such interactions in order to achieve particular pedagogical goals.

L1 writing experience is a useful resource in L2 writing processes, especially for less proficient students or for demanding tasks. Woodall (Reference Woodall2002) observed that less proficient L2 writers switched to their L1s in their think-aloud protocols more frequently than more advanced learners, and that difficult tasks increased the duration of L1 use. Interestingly, an L1 background had a different impact on the quality of L2 texts: for learners from a cognate L1 background, longer periods of L1 use were associated with higher quality L2 texts, whereas for non-cognate background students, the L1 switch was related to lower quality L2 texts. Iwashita & Sekiguchi (Reference Iwashita and Sekiguchi2009), however, argued that L1 effects must be considered in relation to instructional modes. Their analysis revealed that character-based (e.g. Chinese) post-beginners (i.e. students who started with an intensive first-year university course) generally performed better in kanji use and other areas than post-secondary learners (i.e. those who started at secondary school), but that among post-secondary learners, character-based students did not necessarily outperform non-character-based (e.g. English) learners.

Computers can be beneficial in L2 writing, their uses ranging from drafting to revising, but Japanese word-processing input procedures can be a challenge for L2 writers. Chikamatsu (Reference Chikamatsu2003) compared typed and hand-written samples of intermediate Japanese L2 learner text and found that the typed samples outscored the hand-written samples only in word accuracy and kanji use, but not in writing time, nor in the quality of writing measured by t-unit. In addition, the kanji usage ratio improved significantly among mid-skilled, but not among low- or high-skilled students, suggesting that the benefits of computer use vary according to L2 proficiency. Takahashi (Reference Takahashi2009), on the other hand, demonstrated that a computer program encouraging reflective thinking had a positive impact on higher levels of L2 writing processes (e.g. self-evaluation on logical sequence and content), but not on L2 use at lower levels of analytic evaluation. These findings suggest that computer use could promote the transfer of higher levels of thinking from L1 to L2 and ease the burden of character-writing, but that it cannot compensate for learners' limited knowledge in L2 use.

One question that language teachers often ask concerns the types of feedback or error correction that L2 writers find useful. Kubota (Reference Kubota2001) found that, rather than the teacher providing a correct form, an error coding system (indicating the type of an error, such as P for a postpositional particle error) was effective for self-revising in L2 Japanese writing in general. However, the effectiveness of this approach varied depending on error types and L2 proficiency. Particle errors, for example, were relatively easy to fix, partly because of the limited number of particles, but word choice errors were extremely difficult to self-correct. Kubota pointed out that, while high proficiency students might simply need to be informed of the locations of errors as their grammaticality judgments improved and their proficiency increased, lower proficiency students might need more explicit feedback. Teacher feedback, however, differed significantly in rating criteria and bias patterns (e.g. their overall severity for extremely high- or low-ability students), although experienced JFL teachers generally produced consistent self-evaluations (Kondo-Brown Reference Kondo-Brown2002). Furthermore, Hirose (Reference Hirose2004) found that, unlike teacher feedback, peer responses addressed the content of essays more frequently than linguistic structures (e.g. grammar, spelling, mechanics), but that peer comments regarding the content were not incorporated in student revisions at all, suggesting that the effectiveness of feedback was confounded with feedback types and student perceptions.

As we have seen, research in L2 reading and writing addresses key themes in SLA, such as cross-linguistic transfer, L1–L2 interface, cognitive load, socio-cultural factors affecting learning contexts, the role of instruction and feedback, and learner perception and goals. Too much focus on the end products might lead one to overlook insights into L2 Japanese reading and writing processes. A student's failure to incorporate teacher feedback in a revised draft, for instance, might be due to a lack of attention, a misunderstanding, or the inability to correct errors due to limited L2 proficiency. Research in this area therefore requires more in-depth analysis of on-going processes from multiple perspectives.

1.4 Interlanguage pragmatics and language socialization

Research on L2 Japanese has maintained a strong tradition in its investigation of interlanguage pragmatics and language socialization. Topics explored in this area of research include types of speech act, such as requests or refusals, the use of interactional particles and other utterance-final expressions, honorific speech styles, demonstration of listenership and alignment, comprehension of formulaic expressions and indirect meanings, or enhancement of discourse cohesion and coherence. The existence of a rich set of indexical resources that are critical for Japanese interaction, but hard to learn and teach, appears to have prompted many researchers to examine closely L1 and L2 speakers’ language use. Anecdotes and folk beliefs concerning differences between Eastern and Western styles of communication may have also contributed to this development.

One of the traditional approaches in interlanguage pragmatics research is to compare L1 and L2 speakers’ performance in the use of interactional particles (Masuda Reference Masuda2009), hearsay evidentials (K. Ishida Reference Ishida2006), honorific forms (Ikeda Reference Ikeda and Taguchi2009), or particular speech acts and interactional moves. Iwasaki (Reference Iwasaki2009), for instance, compared how L1 and L2 speakers state and support their opinions in an interview setting, and found that L1 speakers frequently used various mitigation devices to seek common ground with the interlocutor, whereas L2 speakers rarely employed such devices. Nakai (Reference Nakai2002), on the other hand, analyzed how L1 and L2 speakers support their co-participants in topic openings and closings in face-to-face conversation. She reported that L2 speakers’ lack of knowledge of topic-shifting devices caused them to open and close topics in a rather abrupt manner; as a result, these L2 speakers created an impression that they did not understand or appreciate the topics introduced by their L1 interlocutors.

Receptive and interpretive skills for certain pragmatic features are yet another area where L1 and L2 speakers have been compared. Hagiwara (Reference Hagiwara and Taguchi2009), for instance, reports that L1 speakers and intermediate L2 learners differ in their interpretation of formulaic utterances. Cook (Reference Cook, Rose and Kasper2001) demonstrates how native-speaking teachers and L2 learners showed different levels of sensitivity towards the inconsistent use of desu/masu style observed in audio-recorded self-introductions. Similarly, H. Ishida (Reference Ishida2006) points out qualitative differences between L1 and L2 speakers’ perception and interpretation of back-channel cues.

The performance of L2 learners on different levels has also been studied in order to trace the general trend of their development. Taguchi (Reference Taguchi2008), for instance, compared the ways that beginning, intermediate, and advanced level learners interpret indirect meaning, and reported that a proficiency effect was observed on accuracy but not on comprehension speed. Further, Watanabe (Reference Watanabe2003) provided a close examination of ACTFL oral proficiency interviews (OPIs) at the Intermediate-High through Superior levels, paying special attention to L2 speakers’ use of cohesive and coherence devices. Her analysis shows that upper-level speakers used a wider range of devices to extend discourse and to signal discourse organization. Likewise, Geyer (Reference Geyer2007a, Reference Geyer2007b) sheds light on the close relationship between pragmatic, grammatical, and discourse competence in learner language by analyzing the use of contrastive markers in OPIs involving Novice to Superior L2 speakers. The results of her study also suggest that upper-level speakers employ various lexical and grammatical markers in an effective manner to foreground their assertions while properly inserting self-qualifications. These studies show how important it is to teach students how to use a range of predicate endings, connective words, discourse markers, and sentence-final and clausal particles to create extended connected discourse.

The identification of gaps between L1 and L2 speakers’ performance has also inspired studies of the effect of explicit instruction. Tateyama (Reference Tateyama, Rose and Kasper2001, Reference Tateyama and Taguchi2009), for instance, examined how consciousness-raising activities, which include viewing video clips and reviewing their transcripts, may facilitate L2 speakers’ understanding of pragmatic routines, such as thanking, apologizing, or requesting. Yoshimi (Reference Yoshimi, Rose and Kasper2001), K. Ishida (Reference Ishida and Taguchi2009), and Kakegawa (Reference Kakegawa and Taguchi2009) investigated how instructional interventions would improve learners' understanding of the use of interactional discourse markers (n desu, n desu kedo, n desu ne), different speech styles, and interactional particles. While these studies demonstrate some positive results from explicit instruction, they also acknowledge the complex factors that contribute to the development of discourse-pragmatic competence. Remaining questions concern the effect of learners' motivation and other affective factors, the applicability of the knowledge gained through short-term explicit instruction to situations outside the classroom and experimental settings, and the relationship between explicit instruction and incidental learning that may occur through learners' exposure to, and socialization in, real-life interaction.

A growing number of qualitative studies explore these questions by examining how L2 speakers participate in classroom and out-of-classroom interactions. Ohta's (Reference Ohta, Rose and Kasper2001a, Reference Ohta2001b) longitudinal classroom research is a seminal example. She documented how novice learners developed an interactional style of listener response not only through the teacher's explicit instruction, but also through the learners' recurring participation in interactional routines in the language classroom. In a different study, M. Ishida (Reference Ishida2006) described how an L2 learner, through repeated participation in decision-making activities with a native speaker, came to understand the functions of modal expressions. The learner also gained the ability to anticipate the trajectory of the decision-making sequence.

The process of language socialization has also been explored in the study-abroad context. M. Ishida (Reference Ishida and Taguchi2009), for instance, investigated how one learner, during his nine-month study abroad experience, expanded his interactional repertoires through the use of the particle ne. Iwasaki (Reference Iwasaki2010) compared five learners' uses of desu/masu and plain forms before and after their study abroad, and showed that they all gained some understanding of the social meanings of style shift, but that they did not all make the same use of this resource. The results indicate that these learners established their ability to make their own choices as to how they wanted to present themselves in different social contexts. Indeed, Cook (Reference Cook2008) offers a rich analysis of dinner-time conversations between learners of Japanese and their host families, revealing how these learners were socialized to use an appropriate speech selection between desu/masu and plain forms, as well as how they accomplished the socialization of their identities through the style shift.

These recent studies call attention to the issue of identities involved in the development of interactional competence. Rather than assuming that there is a set of norms and patterns of language use uniformly adhered to by all speakers of Japanese, they explore how various pragmatic factors come into play to create speakers’ social or situated identities. This line of research contrasts sharply with the studies reviewed earlier in this section, which generally compare faceless, uniform categories of L2 learners as deficient communicators and L1 speakers as target models. In addition, increasing numbers of recent studies (e.g. Ishihara & Tarone Reference Ishihara, Tarone and Taguchi2009; Kumagai & Sato Reference Kumagai and Sato2009) have addressed how learners' experiences of language socialization can differ, depending on what kinds of social communities they want to belong to or what kinds of identities are attributed to them by other people. L2 speakers do not have to follow idealized native speaker models all the time and the mastery of the target models does not necessarily guarantee that they will be treated like natives. To this end, studies such as Fukuda (Reference Fukuda2006), Zimmermann (Reference Zimmerman2007), and Suzuki (Reference Suzuki, Nguyen and Kasper2009) offer microanalyses of the ways in which linguistic or cultural identities are ascribed, claimed, resisted, or co-constructed in interactions between L1 and L2 speakers.

Finally, while traditional approaches tend to focus on linguistic resources, especially those that are most noticeable and considered unique to Japanese, an emerging body of studies investigates how L2 speakers coordinate linguistic and non-linguistic resources in order to accomplish their participation in social interaction (e.g. Mori & Hayashi Reference Mori and Hayashi2006; Okada & Yanagimachi Reference Okada and Yanagimachi2008; Yanagimachi Reference Yanagimachi2009; Mori Reference Mori, Kasper, Nguyen, Yoshimi and Yoshioka2010). The impartial close analysis of video-recorded interactions with no preconceived notions regarding what L2 learners should study can lead to the discovery of previously unnoticed linguistic and non-linguistic resources that enable speakers to accomplish particular social actions. The findings from such research can contribute to the reconsideration of goals and processes of L2 pragmatic development.

1.5 Affective factors in L2 Japanese learning

The study of individual differences in SLA has identified a number of nonlinguistic factors in an attempt to account for why some L2 learners are more successful than others. One possible source of individual differences can be found in affective factors, including motivation, anxiety, attitudes, and learner perceptions. Affective studies are founded on the assumption that students take different approaches to a certain task because they have different goals, feelings, attitudes, and perceptions, which eventually influence their degree of success in L2 learning. Research has demonstrated that learner affect is a complex system of multiple, independent constructs, and that learners' task-specific beliefs can be related to their strategy use. Consequently, affective research has taken at least two distinct methodological directions: one is a large-scale quantitative approach to identify observable factors and explore the relationship between the identified variables and learning behaviors; the other is an in-depth, context-specific analysis focusing on a single or small number of individual learner(s).

1.6 Multilingualism and the acquisition of Japanese as a heritage language

As a result of globalization, an increasing number of school-age children, accompanied by their parents, receive formal education in a foreign country. Many of them struggle to learn the mainstream language as L2 to acculturate to the foreign community in which they are growing up, while also trying to maintain their home language and ethnic identity. Recently, there has been growing interest among L2 Japanese educators and researchers in the language and identity problems faced by multilingual and/or multicultural children. Within Japan, where Japanese is the mainstream language, JSL education for migrant workers’ children has become a major concern (e.g. Kanno Reference Kanno2008). Outside Japan, where Japanese is not the dominant language, the maintenance or acquisition of Japanese as a heritage language (JHL) becomes an issue (e.g. Kondo-Brown Reference Kondo-Brown2006b). While both JSL and JHL learners must receive equal attention, this section provides an overview of issues in multilingualism, mainly from JHL perspectives, with a hope that a focus on Japanese as a minority language illuminates problems associated with multilingualism.

One reason why issues in JHL attract the attention of L2 Japanese researchers is that JHL learners tend to be far more proficient in the mainstream language than in JHL, but their language proficiency demonstrates different traits from that of L2 or first language (L1) (Lee Reference Lee2005; Kondo-Brown Reference Kondo-Brown2006b). JHL learners usually possess strong Japanese listening comprehension skills, acquired mainly through interactions with their Japanese-speaking parent(s) at home, but varying degrees of speaking and literacy skills. Furthermore, young heritage language (HL) learners' dominant language usually shifts from their HL to the mainstream language as they grow up (language shift). Even with some degree of parental or educational support, the majority of school-age Japanese-speaking children attending hoshuukoo (‘supplemental’ schools outside Japan that offer the Japanese school system curriculum) in the United States were seen to become English-dominant partial bilinguals (Kataoka, Koshiyama & Shibata Reference Kataoka, Koshiyama and Shibata2005). Thus, JHL learners' linguistic and cultural competence cannot be equated with that of L1 learners, who usually use the mainstream language only. At the same time, JHL learners demonstrate different language skills, cultural awareness, and motivational orientation from JFL learners, who often choose to learn Japanese despite having little prior linguistic and/or cultural background knowledge before being exposed to the language (Kondo-Brown Reference Kondo-Brown, Dornyei and Schmidt2001a, Reference Kondo-Brown2001b, Reference Kondo-Brown2005; Lee Reference Lee2005). Language instruction for JHL learners should therefore be tailored to their linguistic and cultural needs (Lee Reference Lee2005; Kondo-Brown Reference Kondo-Brown2006b).

In addition, JHL learners show considerable individual variation in their bilingual competence, which begins to broaden around the fourth grade (Kataoka, Koshiyama & Shibata Reference Kataoka, Koshiyama and Shibata2005). JHL learners' bilingual development is substantially influenced by the amount of Japanese used at home (Shibata & Koshiyama Reference Shibata and Koshiyama2001; Kondo-Brown Reference Kondo-Brown2005). Kondo-Brown (Reference Kondo-Brown2005), for example, compared Japanese language proficiency among JFL learners and three subgroups of JHL learners (JHL Descent Group with no Japanese-speaking parent or grandparent; JHL Grandparent Group with at least one Japanese-born grandparent but no Japanese-speaking parent; and JHL Parent Group with at least one Japanese-speaking parent). The JHL Parent Group demonstrated significantly higher grammar knowledge and listening and reading proficiency than the other groups, and made higher self-assessments; and the other two JHL groups did not differ from JFL students.

Researchers in bilingualism are also interested in whether the two language systems develop independently or interdependently. One assumption, known as linguistic interdependence, is that children's bilingual development is founded on a common underlying proficiency (CUP), and that a well-established CUP is transferable between L1 and L2. This position, if correct, predicts an association between high L1 aptitude and high L2 proficiency. Minami (Reference Minami2003), for instance, demonstrated a positive correlation between English and Japanese vocabulary usage among bilingual children (ages 6–12), using Bilingual Verbal Ability Tests (BVAT). Mishima-Mori (Reference Mori2002), however, found no signs of bilingual interaction in two Japanese-English bilingual children's usage of past tense marking and negation. Shibata (Reference Shibata2004) also found no significant relationship between Japanese and English proficiency among JHL college students. These contradictory findings suggest substantial individual variation among JHL learners in their bilingual attainment.

However, adult learners with a high level of JHL literacy show common traits and similar language experiences. First, competent JHL readers had access to sufficient Japanese printed material, in particular ‘light’ reading materials that fostered their interest in recreational reading, in their home and JHL community (Tse Reference Tse2001). Yoshimitsu (Reference Yoshimitsu2000) observed that Australian-born JHL children with permanent residence status who had attained as good command of Japanese as temporarily resident children were motivated to learn Japanese, and enjoyed reading Japanese books and playing with Japanese-speaking friends. Second, successful JHL learners had contact with community institutions that provided literacy experience beyond the home, including religious organizations (Tse Reference Tse2001), JHL schools (Shibata & Koshiyama Reference Shibata and Koshiyama2001) and hoshuukoo (Chinen & Tucker Reference Chinen, Tucker and Kondo-Brown2006).

JHL learners' ethnic identity is another factor that plays a significant role in their bilingual development. In general, Japanese ethnic identity is thought to be associated with a high level of Japanese language proficiency, and participation in a JHL program (e.g. hoshuukoo) promotes Japanese ethnic identity development and eventually has a positive impact on JHL proficiency, particularly on literacy skills (Chinen & Tucker Reference Chinen, Tucker and Kondo-Brown2006). Conversely, ethnic ambivalence or rejection of one's own ethnic identity leads to negative attitudes toward the ethnic group and its language speakers, feelings that extend to a lack of interest in maintaining and developing their HL (Tse Reference Tse2000). These Japanese/non-Japanese continuum models, however, may not accurately describe new generations’ identities as multilingual speakers. McMahill (Reference McMahill2005), for instance, described how a class project (creating a world globe) in a first-grade classroom in a trilingual (English, Portuguese, and Japanese) community school in Japan – a grassroots effort by multilingual parents and teachers to pass on their languages and cultural values to younger generations – created an opportunity for debating the values of non-English or non-Japanese languages and cultures. Such multilingual education evidenced the emergence of children's identities as multilingual speakers. Similarly, Morita (Reference Morita2003), through her analysis of young children's insertion of English personal pronouns into Japanese speech, interpreted Japanese-English bilingual children's code mixing as indicating their emerging identities as bilingual speakers.

An examination of JHL learners' ethnic identities leads to investigations of a broader multilingual context that involves social factors, such as ethnolinguistic vitality and social networks. Sakamoto (Reference Sakamoto2001; Reference Sakamoto and Kondo-Brown2006) examined retrospective data from five Japanese immigrant families living in Toronto, Canada, using Engeström's activity theory framework (1999, cited by Sakamoto Reference Sakamoto2001), and found that Japanese was viewed as a medium for establishing and retaining family bonding, and that English was considered an indispensable socio-economic tool. As a result, all her informants said they supported their children's JHL oral development because good oral communication was key to family bonding, but some parents did not actively enforce the use of written Japanese or honorific discourse because these types of formal Japanese were not necessary at home. Given that individual parents’ efforts and resources are varied and limited, Shibata (Reference Shibata2000) argues that extra-curricular JHL schools/programs are one of the most effective ways for Japanese-born parents to pass on their HL to their children in a location where ethnolinguistic vitality is low. Her case study identified five essential components that would be required to open a JHL school: leadership, parental and community support, teaching methodology, and materials. Siegel (Reference Siegel2004) identified two additional components that were required to assure ethnolinguistic vitality and continuity, namely social networks and administrative policies.

Despite their limitations, a growing number of JHL studies in English-speaking countries address key issues in multilingualism, including learners' ethnographic backgrounds and bilingual competence, their psychological and sociocultural factors, identity issues, L1-L2 relations, and the ethnolinguistic vitality of a multilingual community. These issues should be further investigated, with both conceptually and methodologically rigorous approaches, in different L2 Japanese acquisition contexts involving groups of young learners who have not been examined by the studies reported in this section (e.g. JSL children in Japan and JHL learners in other areas). A review of published works in those areas would be useful for obtaining a better understanding of the problems faced by multilingual children.

2.1 Instructional approaches and curriculum design

During the last two decades, efforts to establish a common standard or framework for language education have been seen in various parts of the world. Both the Standards for Foreign Language Learning: Preparing for the 21st Century (hereafter, the U.S. Standards) and the Common European Framework of Reference for Languages: Language, Teaching, and Assessment (CEFR) first appeared in 1996 and went through revisions during the following decade. More recently, in 2010, the Japan Foundation released the JF Standard for Japanese Language Education (JF Standard, http://jfstandard.jp), which is modeled on the CEFR. These documents all emphasize the necessity of identifying learners' needs and goals, and encourage educators to consider what learners should be able to do in L2 upon completion of a particular unit, as they develop materials, activities, and assessment tools.

The establishment of these standards has had varying degrees of impact depending on the language, level of instruction, institution, and region. Falsgraf (Reference Falsgraf and Nara2001), for instance, discusses the fundamentals of a standard-based system with reference to developments in the state of Oregon, where strong legislative and policy mandates have accelerated JFL educators' adaptation to this new approach. He concludes that the standard-based approach, which lays out clear, specific, performance-based goals for instruction and requires reliable and practical assessment methods to monitor learners' progress towards those goals, has improved not only student performance but also teacher quality. In addition, Yamada & Moeller (Reference Yamada and Moeller2001) reported on their design of a project-based unit around pen-pal letter exchanges, following the guidelines of the U.S. Standards. Their action research study describes how a classroom lesson can be changed from a prescriptive form-focused approach to a descriptive meaning-focused one. A number of similar action research studies have been presented at regional, national, and international conferences during the last decade.

This shift towards an emphasis on meaning and effective communication and away from purely form-focused instruction has promoted a number of instructional methods. While the influence of communicative approaches is still prevalent, others, such as content-based, sociocultural, or literacy-oriented approaches, have recently gained recognition. However, the application of these approaches, first developed for teaching English and other commonly taught European languages, to JFL, raises some unique issues. For instance, as discussed earlier, non-Western orthographic systems present a challenge to a balanced development of oral proficiency and literacy in JFL. Indeed, the timing of the introduction of the Japanese syllabaries (Hatasa Reference Hatasa2002; Dewey Reference Dewey2004a) and the place of handwriting – given the existence of word-processing technologies (Allen Reference Allen2008) – have been subjects of controversy in their own right. Further, for English and other European language speakers, the grammatical structures of Japanese, being different from those of their native language, and the lack of cognates between their L1 and L2, present additional challenges. As a result, ideas based on the instruction of commonly taught European languages cannot be easily or speedily adapted to JFL contexts.

Acknowledging such challenges, many still advocate the importance of content-based instruction (CBI). Spence-Brown (Reference Spence-Brown2001), for instance, argues that by incorporating Japanese studies content into language courses, learners can gain social and cultural knowledge essential for effective communication: they are learning the language, and at the same time, using it to do something meaningful. To overcome obstacles such as the lack of material and teacher expertise, she recommends collaboration between Japanese studies experts and language experts.

Douglas & Kataoka (Reference Douglas and Kataoka2008), however, emphasize the importance of designing systematic and consistent scaffolding activities in order to make selected materials accessible for students whose proficiency may not be advanced enough to tackle them on their own. Similarly, Hasegawa & Kambara (Reference Hasegawa and Kambara2008) argue that, for understanding authentic materials, it is critical to develop students’ ability to appreciate aspects such as rhetorical structure, organizational cues, cohesive devices, which are typically not covered in introductory language courses (see also Fujii Reference Fujii2005 for a similar discussion). Further, Morioka, Takakura & Uchida (Reference Morioka, Takakura and Ushida2008) introduce web-based multilevel materials developed by the California State University system for promoting Japanese CBI, showcasing sample activities from a second year course and a heritage language course. In addition, Chikamatsu's (Reference Chikamatsu2009) case study of CBI implemented in an advanced level university classroom documents how a local community of Japanese and Japanese-Americans were brought into the students’ learning experience.

One of the areas where CBI has been adopted for some time is language for specific purposes (LSP), such as technical Japanese (e.g. Tsutsui Reference Tsutsui and Nara2001) or business Japanese (e.g. Wei Reference Wei2003; Takami Reference Takami, Jourdenais and Springer2005). Wei (Reference Wei2003), for instance, introduces a case study of a course she developed for a group of advanced learners of Japanese, who were legal specialists soon to be assigned to Tokyo. The goal of this course was to develop their understanding of Japanese corporate culture and current affairs, at the same time promoting their language proficiency. Even for such a small, coherent group of learners, Wei found it challenging to locate appropriate course materials and design suitable instructional activities. In a different study, Takami (Reference Takami, Jourdenais and Springer2005) discusses an intermediate level business Japanese course offered for the general undergraduate population. In order to make authentic materials on relevant topics accessible to learners at this level, Takami found it necessary to rewrite some texts, a move that some have questioned. In addition, although the implementation of CBI is not their main focus, Thomson, Masumi-So & Osho (Reference Thomson, Masumi-So and Osho2001) also report on how an intermediate level course was developed centering on the practices of the tourism and hospitality industry, a major employer of Australian non-native speakers of Japanese.

For these newer approaches to succeed, however, it is critical that both instructors and learners understand their benefits. A survey of Japanese, Spanish, and Swahili students conducted by Kubota, Austin & Saito-Abbot (Reference Kubota, Austin and Saito-Abbott2003), for instance, reveals that while the authors consider it important to include sociopolitical issues in foreign language classrooms, students often perceive language learning and content learning as separate matters and tend to prefer language courses to focus on grammar, vocabulary, and basic communication. Such perceptions about the nature of language courses may be the results of students’ socialization through exposure to classroom interactions, textbooks, and public discourse that represent the persistent influence of conventional views of language learning. Parry's (Reference Parry2000) review of introductory Japanese language textbooks, for instance, suggests that despite the authors’ claims that their textbooks adopt a Communicative Language Teaching (CLT) approach, the approach is only partially reflected in the design of exercises. According to Parry, while meaningless form-concerned grammatical exercises are not very common, a large portion of the exercises still impose control over the expected correct utterance. Mori's (Reference Mori2005) analysis of the sample dialogs included in introductory textbooks also provokes a similar observation. The use of these introductory textbooks may unintentionally cultivate a particular view of what classroom language learning should be like, although it must also be acknowledged that the value of these traditional teaching and learning practices cannot be easily dismissed.

In order to compensate for the limitation of textbooks and classroom instruction, several case studies report web-based projects designed to supplement regular classroom activities. Fukai, Nazikian & Sato (Reference Sato and Doerr2008), for instance, describe how the authors introduced sociocultural approaches to assessment activities by incorporating web-based peer learning and portfolio projects. These projects provided students with opportunities to engage in self- as well as peer-assessments and reflect on their own language learning experiences. Ishihara (Reference Ishihara2007) introduces a web-based pragmatics curriculum developed for JFL learners, which aims to raise learners' awareness. The web-based self-study curriculum encourages students to take the perspectives of researchers or ethnographers, and examine language use in a descriptive manner.

The trials and errors documented by the studies reviewed in this section offer language educators invaluable information that could help them design lessons and curricula suitable for a particular local context. Obviously, there is no one-size-fits-all solution to issues such as what kinds of content should be selected or how to balance form-focused instruction, contents, tasks, and activities, and the increasing diversity of learners that have to be accommodated in the same program may exacerbate these difficulties. For instance, while the challenges in learning Japanese experienced by speakers of English and other European language were mentioned earlier in this section, the situation is dramatically different for Chinese and Korean learners. Their knowledge of characters gives Chinese learners of Japanese an advantage over others who have to start learning them from scratch. Because Japanese and Korean grammatical structures are very similar, Korean learners of Japanese tend to develop their oral proficiency much faster than speakers of other languages. However, with the globalization of higher education, more and more classrooms have to accommodate students from diverse language and cultural backgrounds learning together, not only in Japan, but also elsewhere. Students differ not only in their language backgrounds, but in their interests and career aspirations, and this also makes it difficult to select a topic or activity that will serve everyone involved. To this end, the ‘can do’ list and the portfolio-based assessment introduced by CEFR and JF Standard may help learners and teachers identify the trajectory of each learner, as well as common features among a group of learners. Furthermore, as we shall discuss in the next section, while the U.S. Standards have been influential in the development of language education in the U.S. over the past ten years, they have not escaped criticism. Studies in the next decade, we hope, will continue to address how these standard-setting documents should influence language teaching practices, as well as how they should be modified by reflecting the needs of the rapidly changing world in which language learners find themselves.

2.2 Conceptualization of ‘culture’ in language education

Learning a foreign language cannot be accomplished without learning the culture embodied by, and constructed through, the language. While this view is widely accepted among educators and researchers, the complexity and elusiveness of culture results in diverse perspectives and approaches as to how culture should be conceptualized and integrated into foreign language education.

Kubota (Reference Kubota2003) discusses the danger in the U.S. Standards’ tendency to reinforce static, fixed images of a given culture as a representation of a homogeneous group. Although it is an important step toward developing an understanding of the differences and similarities between the Self and the Other, Kubota warns that prescriptive approaches that underscore binary judgments of what is ‘correct versus incorrect’ about the culture may prevent students from developing a deeper and more nuanced understanding. Instead, Kubota argues for the importance of promoting students’ critical thinking or, in other words, a descriptive understanding of a culture with the appreciation of diversity within the culture as well as its dynamic and discursive construction. This alternative perspective has been observed in a number of studies published in the last decade (e.g. Tai Reference Tai2003; Kubota Reference Kubota, Mori and Ohta2008; Sato & Doerr Reference Sato and Doerr2008; but see Makino Reference Makino, Sato and Doerr2009 for a different view).

This critical approach examines how ideologies are introduced and recreated in Japanese language classrooms and language textbooks. Matsumoto & Okamoto (Reference Matsumoto and Okamoto2003), for instance, reviewed five commonly used textbooks and showed how they provide simplified accounts of sociolinguistic phenomena such as communication styles (politeness, indirectness, self-effacement), speech styles (formal versus informal), and ‘standard’ Japanese and regional dialects. They caution that such simplification and over-emphasis on generalized cultural values can actually hinder learners' development of sociolinguistic competence. Indeed, Shimizu's (Reference Shimizu and Taguchi2009) study of JFL and JSL learners' responses to compliments attests to the unwanted consequence of simplified and generalized instruction about culture. Specifically, the results of his study indicate that JFL learners frequently responded to a compliment with an explicit denial, which textbooks often introduce as a generalized ideal response.

Likewise, Heinrich (Reference Heinrich2005) points out that the language ideologies that surface in JFL teaching materials developed in the 1970s and 1980s often legitimize socially accepted knowledge originating in the educated middle class of urban centers. Ando (Reference Ando2005) also argues that the concept of ‘correctness’ prevalent in Japanese language classrooms is heavily influenced by the ideology of standard Japanese. She encourages Japanese language instructors to reflect on language used in real-life situations and reconsider what counts as ‘correct’ and how, and by whom, that correctness has been determined. Furthermore, Siegal & Okamoto (Reference Siegal and Okamoto2003) and Thomson & Otsuji (Reference Thomson and Otsuji2003) examine how textbooks tend to introduce stereotyped, hegemonic images of gender roles and gendered speech styles, and suggest that the introduction of traditional norms and viewpoints should be accompanied by critical reflections on how they were established and how they may or may not apply to real-life situations.

As mentioned earlier, in order to curtail the tendency to introduce culture as a set of simple, unambiguous norms, Spence-Brown (Reference Spence-Brown2001) advocates the effective integration of language instruction and Japanese studies and CBI; Ohara, Saft, & Crookes's (Reference Ohara, Saft and Crookes2001) attempt to incorporate a feminist critical pedagogy into a beginning level course, or Iwasaki & Kumagai's (Reference Iwasaki and Kumagai2008) action research on critical reading implemented in an advanced level reading course provide examples of such integration. In order to pursue opportunities of collaboration with Japanese Studies experts, as proposed by Spence-Brown, language experts need to expand their understanding of how culture is approached in different disciplines within the humanities and beyond. Indeed, many students in higher education have exposure to courses in anthropology, history, literature, political science, or sociology, and the knowledge and critical perspectives that they gain through these courses can also inform their learning in the language classroom.

Learners' exposure to culture, of course, is not limited to the formal instructions discussed so far. The study by Shimizu (Reference Shimizu and Taguchi2009), mentioned earlier, also demonstrated that JSL learners who had ample opportunities to interact with native speakers outside the classroom appeared to have modified their understanding of how one should respond to compliments, which was originally based on textbook instruction. Itakura's (Reference Itakura2004) study of a collaborative intercultural email project between Hong Kong learners and native speakers of Japanese also explored how the learners form, reaffirm, or modify cultural stereotypes about Japan by interpreting various types of input, including classroom teaching and remarks from native speakers and fellow learners, as well as the mass media. The effects of learners' interactions with native speakers upon the formation of their cultural understanding is also evident in Cook’ (Reference Cook, DuFon and Churchill2006) and Iino's (Reference Iino, DuFon and Churchill2006) studies, which examine interactions between learners studying abroad in Japan and their host family members. Their everyday interactions provided both parties with opportunities to become aware of each other's cultural assumptions, and to jointly construct or renegotiate folk beliefs.

Finally, the recent surge of interest in Japanese popular culture cannot be dismissed. Increased access to popular culture materials, such as manga, anime, dramas, games, and music, and the rapid growth of online fan communities have certainly changed and expanded the environments in which the learning of Japanese language and culture takes place, as observed by Thorne (Reference Thorne2009). While this observation has been confirmed by numerous anecdotes shared among Japanese language educators, until now few systematic studies have been conducted to investigate the impact of popular culture fandom upon the learning of language and culture. Fukunaga (Reference Fukunaga2006) is one of a few published studies that explore how students who self-identify as anime fans incorporate their anime knowledge while learning the language, and how they apply newly acquired knowledge of vocabulary, grammar, writing systems, and culture through their language classroom experience to enhance their appreciation of anime.

While restricted or simplified presentations of culture in language classrooms have long been subject to criticism, a wide range of cultural products, practices, and perspectives have become easily available to learners outside the classroom. Nevertheless, these learners may not have a sophisticated lens through which to appreciate the discursive construction of culture. Questions of how to conceptualize culture and how to introduce it in language classrooms need to be reconsidered vis-à-vis ever-changing environments, including reality and cyberspace, where today's learners are situated.

2.3 Classroom discourse

Along with the exploration of innovative instructional approaches, the investigation of classroom discourse has also expanded its scope during the last decade. While classroom research used to focus on teacher-student interactions, a growing number of recent studies have examined how peer-to-peer interactions are carried out in actual classroom settings. In the meantime, the methodological and conceptual frameworks adopted for classroom research have also become diversified.

Studies conducted within cognitive/psycholinguistic frameworks, for instance, typically measure language learning in terms of the expansion of knowledge of linguistic systems and their accurate execution. One of the key questions explored in this tradition concerns what kind of interactional moves effectively promote learners' recognition of their errors and their eventual production of correct forms. Iwashita (Reference Iwashita2003) examined how negative feedback and positive evidence provided in task-based interaction affect learners' acquisition of two target grammatical structures. The results indicate that interactional moves containing positive evidence were beneficial only for learners who received an above-average score on the pretest, whereas implicit negative feedback, especially recasts, contributed to short-term development of the target grammar regardless of learners' pretest scores. Similarly, Ishida (Reference Ishida2004) studied the effects of recasts on the acquisition of a grammatical form and found that the number of recasts provided during the treatment period was significantly correlated with the students’ overall accuracy, as measured both in the immediate and delayed post-test periods. In addition, Lyster & Mori's (Reference Lyster and Mori2006) comparative analysis of immersion programs, one in French, the other in Japanese, revealed Japanese learners' more frequent and more accurate responses to teachers’ corrective feedback. The researchers suggest that the results were due to the differences in commonly employed class activities between the two settings: the Japanese program is described as focusing more on linguistic accuracy than on the contents conveyed. The typological difference between the target languages and the students’ native language is also discussed as a possible cause for the different tendencies in the two programs.

Wei (Reference Wei2002) specifically examined how learners perceive recasts provided by their teachers in the classroom. Based on stimulated recalls concerning the relevant segments of video-recorded classroom interactions, she reported that the learners, in general, were able to understand the feedback given on their morphosyntactic, phonological, and lexical errors, but that there were also a notable number of occasions when the intended targets of correction were misunderstood. Learners' perception of corrective feedback was also studied by Yoshida (Reference Yoshida2008), who explored corrective feedback provided by peers during pair work. She concluded that learners' responses to corrective feedback may indicate that they have noticed the correction, but may not always suggest their understanding of errors. Yoshida also suggested that the effectiveness of corrective feedback in pair work may be influenced by the level of the learners' satisfaction with their assigned pair work roles.

The last decade has also seen a surge of interest in sociolinguistic and sociocultural approaches to the study of classroom discourse. These approaches conceptualize L2 learning as the increased participation in L2 social interaction rather than the expansion of knowledge concerning L2 linguistic systems, and critically review the analytical procedures commonly adopted in earlier studies. Ohta's study (Reference Ohta, Hall and Verlpaetse2000, Reference Ohta2001b) of students’ private speech, for instance, demonstrated the importance of paying attention to how teachers’ recasts and other corrective feedback might affect the learning of students other than those to whom the feedback was addressed. Likewise, Ohta (Reference Ohta2005), as well as Foster & Ohta (Reference Foster and Ohta2005), illustrated the limitations of coding and counting methods typically adopted by studies on corrective feedback. For example, moves that have been classified into categories, such as comprehension checks, confirmation checks, or clarification requests based on their forms, may actually serve other functions. Furthermore, these studies argue that moves that have not received much attention in the past, such as co-construction and prompting, appear to play a critical role in assisting other learners to accomplish the tasks at hand. Using conversation analysis, Mori (Reference Mori, Gardner and Wagner2004a) also demonstrated the importance of analyzing interactions involving language learners not only from the perspective of error correction and information transfer, but also from the perspective of the achievement of social actions and sequential development.

Pair or group work taking place in the language classroom is another area that has received increased attention during the last decade. Iwashita's study (Reference Iwashita2001), conducted within the cognitive paradigm, compared the quality of peer interactions in mixed level dyads and same level dyads based on the number of interactional moves and modified output made. Ohta's qualitative analyses (Reference Ohta2001b, Reference Ohta and Nara2001c), informed by Vygotskian sociocultural theory, revealed how peer interactions provide different types of opportunity for learners' participation from teacher-fronted interactions and how learners accomplish peer assistance not only by correcting errors, but also by offering assistance for the partner's performance. Continuing this line of investigation, Ohta (Reference Ohta, Mori and Ohta2008) examined laughter observed in peer interaction in the classroom as an indication of the learners' social interactive, affective, and cognitive states. Bushnell (Reference Bushnell2008) also analyzed how learners in the classroom engage in language play, and thereby participate in the teacher-assisted pedagogical activities on their own terms. He argued that language play provides opportunities for further development of sociolinguistic competence, as well as for memorable encoding of the target language.

While Ohta & Bushnell argued for the potential benefit of laughter and language play on learners' linguistic development, Yoshida's case study (Reference Yoshida2007) reported that a student's self-expressive jokes or unique answers were often considered distracting by his instructor. This indicates a mismatch between the instructor's and the student's expectations about the goal of classroom interaction. The gap between teachers’ intended outcomes and students’ actual performance was also demonstrated by J. Mori (Reference Mori2002). Her study illustrates an inadvertent effect of task instruction on the sequential development of talk between two learners and a guest speaker. The influence of task instructions and task prompts on the quality of resulting peer interaction is also examined by Mori (Reference Mori2004b) and Mori & Hasegawa (Reference Mori and Hasegawa2009). Furthermore, these studies demonstrate that even when engaging in the same activity, learners demonstrate their convergent or divergent orientation towards different aspects of language learning, including pronunciation, grammar, or discourse, which may not match the instructors’ or researchers’ expectations.

Other studies of classroom discourse address more directly how a particular instructional approach can be implemented and what kinds of learning opportunities may be realized as a result. DiNitto (Reference DiNitto2000), for instance, explored how two student groups responded differently to a task instruction aimed at enhancing collaborative learning. Her study suggested that for collaborative learning to yield positive results, it is essential to establish a classroom culture that incorporates the students’ socialization into the particular learning style. Yoshimi (Reference Yoshimi, Mori and Ohta2008) documented the effect of lessons designed to explicitly encourage students to draw upon their experiences, intuitions, and knowledge in the classroom for their development of interactional competence. Furthermore, Tateyama & Kasper (Reference Tateyama, Kasper, Soler and Martínez-Flor2008) reported how the presence of a native-speaking guest speaker in the language classroom gives opportunities for different participation frameworks for pragmatic development. Kumagai (Reference Kumagai2007) examined how conflicts between an instructor's teaching agenda and students’ learning agendas may arise in the language classroom. Her critical discourse analysis suggests how such moments of tension could offer rich opportunities to foster students’ critical literacy.

As innovative instructional approaches continue to be introduced into the classroom, rigorous classroom research, informed by solid methodological and conceptual frameworks, remain critical for timely evaluation of how these approaches change the quality of learners' experiences. The studies published so far tend to examine classrooms in higher education largely because most researchers and readers of these studies are affiliated with higher education. Classroom research should be conducted in a wider range of settings to address such issues as the articulation between programs or the development of effective LSP courses.

2.4 Study abroad

As mentioned in the previous sections, an increasing number of studies have explored how language learning takes place outside the classroom, and how learning inside and outside the classroom can influence each other. In particular, the impact of the study abroad experience upon learners' development has, in recent years, attracted the interests of researchers, educators, and administrators, as the emphasis on the internationalization of higher education continues to accelerate. As a result, a number of conference panels and symposia on this topic have been organized during the last decade.

One such symposium yielded a special issue of Japanese Language and Literature offering an overview of study abroad related research (Dewey Reference Dewey2007), conceptual frameworks to understand or enhance learners' study abroad experiences (Kawamura Reference Kawamura2007; Noda Reference Noda2007; Thompson Reference Thompson2007), and detailed accounts of existing programs (Kawamura Reference Kawamura2007; Warnick Reference Warnick2007). However, the number of empirical studies that systematically investigate the effects of study abroad experiences is still limited. Among the articles featured in the special issue, only Iwasaki's (Reference Iwasaki2007) includes an explicit discussion of research methods. This study demonstrates the diversity among four study abroad participants’ goals, expectations, experiences, and gains in language proficiency. It also acknowledges the difficulty of measuring the different types of gains accomplished by the participants through the use of standardized tests, such as the Japanese Language Proficiency Test (JLPT) or the ACTFL Oral Proficiency Interview (OPI).

An approach often adopted by researchers to evaluate gains from study abroad experiences is to focus on specific aspects of learners' proficiency. For instance, Asada (Reference Asada2006) examined how four Chinese students who had received formal L2 instruction in China changed their speech style when exposed to the informal language spoken in daily life in Japan. By focusing on the use of the aspect marker and its reduced form, Asada revealed how the use of the reduced form by the four students steadily increased over time, reflecting the expansion of their social network and their accommodation of native speaking interlocutors’ speech variety (see also Ishida Reference Ishida and Taguchi2009 and Iwasaki Reference Iwasaki2010, discussed earlier, for similar findings). Shimizu (Reference Shimizu and Taguchi2009) compared learners who studied Japanese in their home countries to those who spent an extended period of time in Japan by using discourse completion tasks (DCTs) that elicited their performance concerning compliment responses, and reported that the time spend abroad can trigger the learners' pragmatic development. Dewey (Reference Dewey2004b) investigated the reading development of learners who participated in a summer study abroad program and those who participated in an intensive domestic immersion program. The results of his study showed no significant differences between the two groups with regard to gains in reading comprehension measured by free-recall and vocabulary knowledge, but the study abroad participants’ self-assessment of their reading comprehension was higher than that of the domestic immersion program participants. Dewey also notes the degree of variability in the study abroad experiences, as opposed to the relative homogeneity of the domestic immersion experiences, which might have influenced the results.

Needless to say, the study abroad experience encompasses different types of learning opportunities. McMeekin (Reference McMeekin, DuFon and Churchill2006), for instance, compared classroom interactions with those that study abroad participants experience with host families, paying special attention to exposure to comprehensible input, opportunities for output modification, and focus on form. The results indicated that students overall engaged in more negotiations with the host family to attain comprehensible input than in the classroom, but that the classroom provided more opportunities for student modification of output and negotiations focusing on form. Based on these results, she concluded that a combination of in- and out-of-class interaction with native speakers maximizes the benefits of the study abroad experience. In addition to differences arising from the settings of language use, study abroad participants’ racial or ethnic backgrounds can affect the quality of their experience in Japan (Kumagai & Sato Reference Kumagai and Sato2009).

For the enhancement of effective articulation of pre-departure, study-abroad, and post-study-abroad training, the understanding of the diverse experiences participants encounter during their time abroad is invaluable. Kurasawa & Nagatomi (Reference Kurasawa and Nagatomi2006) presented an example of such an attempt by examining how the Massachusetts Institute of Technology evaluated the experiences of participants of its internship program in Japan and incorporated the results into the restructuring of curricula at the home institution. While similar small-scale action research projects appear to have been undertaken at various institutions, further research is still necessary to uncover the complex web of learning opportunities afforded by study abroad and their effects upon L2 development. In addition, an emerging body of research on study abroad experiences emphasizes the importance of examining learners' subjectivity (Kinginger Reference Kinginger2009) in order to understand their experiences abroad. This is another area where study-abroad research conducted in JFL/JSL contexts should continue to expand its scope.