‘Changes’ to the new Japanese-Language Proficiency Test:
Newly emerged language policies for non-Japanese and Japanese citizens
Volume 19, Issue 3 (Article 8 in 2019). First published in ejcjs on 24 December 2019.
In 2010, the operators of the Japanese-Language Proficiency Test (JLPT), the Japan Foundation and Japan Educational Exchanges and Services, introduced a ‘new’ version of the test, claiming to focus on communicative abilities. However, this is contradictory, as the test continues to measure the same components as the previous version—vocabulary, kanji, grammar, reading and listening—despite any addition of speaking or written components. Focusing on the ideologies of kokugo (for Japanese people) as opposed to nihongo (for foreigners) and tracing back through the history of Japanese-language testing and how it is intertwined with language policies, this paper argues the limitations surrounding the new version of the JLPT, which is the result of a lack of integration of Japanese testing instruments to actual language testing models from literacy to modern testing models, as well as reduced information on its procedures. This is further connects into newly emerged language policies that arose after the introduction of the new JLPT. Details of these policies and their implications are discussed.
Keywords: Japanese-language proficiency test, language policies, testing, kokugo, nihongo.
The Japanese-Language Proficiency Test (JLPT) is the largest Japanese test designed to measure the proficiency of examinees in the Japanese language. Two jointly operated organisations run the test: the Japan Foundation, backed by the Ministry of Foreign Affairs, and the Japanese Educational Exchanges and Services (JEES), supported by the Ministry of Education, Culture, Sports, Science and Technology (hereafter MEXT1). The test has expanded since its introduction in 1984. It began with approximately 7000 examinees from 15 countries. During the July and December sittings in 2017, the total number of applicants exceeded one million worldwide (JLPT n.d.a). Originally developed by a group consisting of representatives from the Japanese Ministry of Foreign Affairs, the Ministry of Education, the Japan Foundation and the Association of International Education, the test was created due to the need for a standardised Japanese-language test for both inside and outside Japan (Nabei & Busch 1999). The official JLPT website states that the JLPT certification is used to evaluate and certify Japanese-language proficiency for non-native speakers, for employment, pay increases and promotion and forms of qualification (JLPT n.d.b). In 2010, the JLPT’s operators introduced the ‘new JLPT’, claiming to be focused on communicative abilities (JLPT n.d.b), despite any inclusion of speaking or writing sections in the test.
When the original JLPT was created, there was a restriction on those able to take the test; only non-native speakers of Japanese were allowed to sit the JLPT. This was determined by citizenship, as Japanese citizens were not able to undertake it. This aligns with the long-standing ideology, or notion, of kokugo (for Japanese people) versus nihongo (for foreigners). Nihongo has typically been reserved for the Japanese language taught to foreigners versus kokugo, in which Japanese students take kokugo classes with kokugo textbooks (Gottlieb 2005). These terms will be explained later in this paper. The distinction between the names was reflected in the JLPT when it was created. For example, test content specifications for the original JLPT labelled its contents ‘Gaikokujin Nihongo Noryoku Shiken’ no ‘Nintei kijun’ (Foreigner Japanese-Language Proficiency Test’s Accreditation Criteria) (Japan Foundation & Association of International Education 1994, p. 3).
After the introduction of the new JLPT in 2010, scarce academic literature has illuminated exactly what has changed about the test. Shohamy (2006, p. 109) explained that ‘language tests are often introduced with an ulterior motive connected with furthering the policy agendas of those in positions of influence.’ This paper argues that, in this case, Japan also has its language tests interconnected with educational, political and economic factors which must be taken into account to understand the changes since 2010. The changes to the JLPT involve a close connection to newly introduced language policies in Japan. Tollefson (1991, p. 16) explains that language plan refers to ‘all conscious efforts to affect the structure or function of language varieties’, and language policy as ‘language planning by governments.’ This paper contributes to the area of Japanese testing and language policies. It also outlines the current progression of Japanese-language testing.
This article comprises the following sections. First, definitions and explanations of kokugo and nihongo, are introduced. Second, this paper presents a brief historical overview of Japanese-language policies and its relation to Japanese testing, particularly since the Second World War. Third, this paper introduces the original JLPT and the new JLPT and argues the lack of progression in the tests. Fourth, this paper analyses the rhetoric surrounding the JLPT in relation to its examinees. The argument illuminates the political tensions and contradictions surrounding the new JLPT, Japanese testing and recent language policies in Japan; this ultimately connects back to the divide between kokugo and nihongo, the backbone of nihonjinron discourse.
Kokugo and nihongo
Two significant terms are kokugo and nihongo, which are two words for the Japanese language. Kokugo (national language or language of our country) refers to the Japanese language for Japanese people and nihongo (language of Japan) refers to the language taught for non-Japanese people (Gottlieb 2005). Gottlieb (2005) explains how, in the case of English, distinguishing between kokugo and nihongo, between native and non-native speakers, can be achieved by adding extra words; native speaker students attend English classes at school while non-native speakers study ESL or EFL. However, the two terms are much more than mere labels; the difference between nihongo and kokugo has been central to the nihonjinron discourse. A central component of nihonjinron is that Japan is a linguistically homogenous nation, and the Japanese language itself is so uniquely difficult even for the Japanese themselves, let along for foreigners to learn; hence in this regard, race, language, and culture are inseparable to each other (Gottlieb 2005, p. 4). Sugimoto (1999) puts forward the following:
Nihonjinron defines the Japanese in racial terms with nihonjin comprising most members of the Yamato race and excludes, for example, indigenous Ainus and Okinawans as groups who are administratively Japanese, but not ‘genuinely’ so. Furthermore, when Nihonjinron analysts refer to Japanese culture, they almost invariably mean Japanese ethnic culture and imply that the racially defined Japanese are its sole owners. (p. 82)
Sugimoto (1999) considers nihonjinron an equation in which nationality, ethnicity, and culture are dimensions used synonymously, despite a growing number of individuals in Japanese society who do not satisfy the equation due to the globalisation of the Japanese economy. In this sense, nihonjinron asserted that Japanese were one minzoku, and minzoku may mean race, ethnicity, or nation; all of which is tied by that Japanese blood (Befu 2001; Yoshino 2005). This ideology therefore meant that if Japanese people were connected by blood, those without this connection could not be considered Japanese. However, there is clear evidence of diversity in Japanese society. It is estimated that the total minority population in Japan lies between 5.78 million and 7.85 million, which account for approximately 5-6% of the total population (Okano & Tsuneyoshi, 2011, p. 6; Sugimoto 2010, p.7, 191-5 cited in Okano, 2017, p. 154). Amongst this minority population, Yamashiro (2013) points out there are a number of minority groups, or perceived subset groups, residing in Japan: Ainu, Okinawans, Burakumin, ethnic Koreans, foreign workers, Japanese Brazilians, and mixed race Japanese. Yamashiro also describes a subgroup perceived as different from mainstream Japanese is ‘returnee’ Japanese (kikoushijo), those who have been partly raised abroad—often due to business or other reasons—and, upon returning to Japan, are typically less familiar with the Japanese language and social norms. As ethnic and cultural borders manifested through the nihonjinron framework are increasingly challenged, so, too, is the concept of kokugo for Japanese people versus nihongo for foreigners.
The insider-outsider tenets of kokugo and nihongo noticeably apply to kokugo education and nihongo education, as the two have often been perceived as separate entities by political bodies. Nihongo was spread by the Japan Foundation from the 1970s during the period of Japan’s economic development. The purpose was to promote the study of Japan globally (although kokugo was taught in pre-war colonies such as Korea and Taiwan) (Gottlieb 2005). This is strongly related to the matter of ownership; it is argued that ‘the use of kokugo indicates that its remains firmly in Japanese hands’ in terms of ‘ownership’ of the language (Gottlieb 2005, p.15). Such insider-outside tenets of kokugo for Japanese people versus nihongo for foreigners was clearly reflected in the JLPT when it was first created, restricting examinees only to those of foreign citizenship. This is hardly surprising given that the operators of the JLPT were political groups.
However, recently there have been changes blurring the distinction between kokugo and nihongo, as the Association of Kokugogaku changed its name to the Association for Nihongogaku in 2004 (Gottlieb 2005). Academics also argue a view diminishing such difference between the two terms. For example, Nezu, Handa, and Hirata (2009, cited in Handa 2011) assert that nihongo education and kokugo education should not be considered on two completely separate tracks. Instead, the starting point of nihongo education is one that gradually merges in the same direction and reaches the level of kokugo education.
Yet despite the diminishing divide between kokugo and nihongo, politicians continue to use discourse aligning with nationalistic ideologies, perceiving Japan as an homogenous nation (tan-itsu minzoku) with one race, culture, and language (Burgess 2012). This has been reflected on a national level in Japan’s language policies by the Japanese government. Heinrich (2011) describes Japan’s record for language and cultures as dismal, as educational policies for linguistic and cultural minorities are yet to be addressed. However, since the Ainu Cultural Promotion Act in 1997, Ainu has been recognised as a language of Japan, which is an example of the misconception of Japan as a monolingual state changing. The ideology of Japanese as a second or foreign language is geared towards Japanese for ‘foreigners overseas’ rather than Japanese as a second language within Japan, despite the existence of older minority groups (Nakane et al. 2015). Likewise, Japan’s national policies for accommodating cultural diversity are yet to be articulated in the student population (Okano 2012, 2017). Academic literature demonstrates that no action has been taken by the national government despite the increasing number of migrant children who require Japanese language assistance in compulsory schooling, which upholds the myth that Japan is a culturally and linguistically homogenous nation.
Japanese-language policies and testing
The first baseline of literacy: Kana
To understand the progression of Japanese-language testing, Japan’s history of language policies must be considered. This allows an understanding of how its relation to Japanese-language testing has been influenced up until today. At the core of Japanese testing, as an orthographic language, it is unsurprising that written scripts play a vital role in determining language proficiency. This is also reflected in language policies set by MEXT. The Japanese language is written using three types of symbols, each with its own function: Kanji is adopted from Chinese characters and is used for conceptual words and names. Kana are phonetic symbols that have two groups (Hadamitzky & Spahn 1996). These are hiragana, which are used at the end of conceptual words written in kanji and other words not written in kanji, and katakana, which is mainly used for words of foreign origin (Hadamitzky & Spahn 1996). Furigana is also used when necessary. Furigana (or ‘annotating kana’) are small kana written with words in other scripts to provide additional information about the words, including their sounds and meanings (Taylor & Taylor 2014). However, what is deemed a functional level of Japanese, particularly regarding written script proficiency, is not straightforward, as can be observed from the most basic form of language testing: literacy.
Japan has prided itself on extremely high literacy rates—close to 100%—despite these figures being debated among academics (Brown 1991; Carroll 2001; Gottlieb 1995, 2008, 2011; Miller 1982; Unger 1996). Miller (1982, p. 186) explains that no-one knows the exact literacy rate in Japan, despite Japanese educational and official policymakers claiming that ‘94.7 percent of the Japanese population is literate.’ Similarly, Gottlieb (2008) argues that, although the United Nations Development Office Human Development Report for 2003 assigned 99% of the adult population in Japan as literate, this figure is erroneous, as it does not consider those with learning difficulties or disabilities for reading or writing, or migrants yet to reach a functional level of the language. These high figures seem more closely parallel to school attendance rates (Horiuchi 1983, cited in Carroll 2001), rather than those who are functionally literate. This is not to infer that literacy rates are not high, especially considering the educated society and extensive publishing industry reflected in Japan (Gottlieb 2008; Miller 1982). However, an examination of figures prior to Japan’s debatably high literacy rates shows that this has not always been the case.
For example, from the early 1900s until 1912, the literacy levels measure from male Japanese military recruits were clearly varied (see Unger 1996, pp. 32-33). In another study, Rubinger (1992 cited in Unger 1996, p. 33) looked into literacy in Japan starting in 1899 and described details of literacy tests’ criteria, of which criterion for ‘elementary school reading equivalence’ were measured by reading either an elementary school textbook or a simple sentence or short poem in kana (i.e., hiragana or katakana) and a few simple Chinese characters (or kanji). At this time, one can interpret that knowledge of kanji to a certain extent was a necessary component of literacy that was measured.
In Japan, students are required to undergo nine years of compulsory education: six years of primary school and three years of middle school. Most students continue to three years of senior high school. Each year level has prescribed national curriculum guidelines for kokugo, or the national language, by MEXT. Students are required to learn hiragana and katakana in year one and two and they also begin to learn kanji from first year, with a different assigned set for each year level until the sixth year. These kanji students learn throughout primary school are called ‘the education kanji’ or kyōiku kanji. Policymakers chose 1006 characters based on simplicity, relevance to function in everyday life and ability to form compound words (Gottlieb 2008, p. 38). The remaining kanji to be learnt throughout middle school are called jōyō kanji or, more recently, shin-jōyō kanji, which was altered in 2009 due to reform changes. These alterations were made by policymakers, but with unclear justifications of the reform based on statistical language usage (Watanabe 2015). MEXT’s perception of the most basic level of ‘functional’ Japanese is far simpler than the number of kanji to be learnt throughout compulsory schooling.
After the Second World War, only an initial couple of surveys showed signs of variability in literacy rates in Japan before the disputable near 100% figures came into place within a rather short period of time and remains to this day. Such surveys include those conducted by the Civil Information and Education Section in 1948 and another smaller survey from 1955 and 1956 (see Unger 1996 pp.36-37). Ever since these surveys, similar evidence of variability in literacy rates is essentially non-existent. The debate of unrealistically high figures of nearly 100% were soon to follow in just a few years, and no surveys of literacy would demonstrate similar details, or even estimates, of those deemed functionally literate, to date. For example, in the early 1960s, when discussing the results of education and script reforms, the education ministry’s extremely high figures of 99.5% literacy were disputed (Gottlieb 1995, pp. 159-160). Simply high figures of literacy without variation can be found after the 1960s from government officials, leaving the issue of exactly how literacy was and has been measured over the past number of decades unclear and questionable.
Another issue contributing to high literacy figures is the definition of literacy in Japan. Brown (1991) argues that the education ministry considers anyone able to comprehend at least one of the three written scripts ‘literate.’ Therefore, as essentially nothing is written in one script, one could meet this standard of being literate despite not being able to read anything. This means that the baseline of meeting the bare minimum of one written script (simply hiragana) implies that kanji is not taken into account when literacy rates are measured. This is unlike the literacy surveys from the early 1900s. Looking back on the Ministry of Education’s national curriculum guidelines, this means that, at the most basic level, people are considered literate if they have at least knowledge of hiragana, which is learnt throughout year one and two of primary school. Therefore, by the end of year two and the beginning of year three, students ought to be literate, as it is assumed they have at least learnt hiragana. However, there are other different functional levels of the Japanese language that are set by MEXT, which is the number of kanji one is supposed to learn. This has explained the most basic form of literacy, which shall be referred to as the first baseline of literacy (i.e., kana) throughout this paper.
The second baseline of literacy: Kyōiku kanji
With as many as 3000 to 3500 kanji characters in use in newspapers and advertisements (Seeley 1991), kanji is important to being able to read written Japanese. As such, a certain level of kanji use is perceived by the Ministry of Education as apparent in terms of reading and writing. In pre-war Japan, 1360 kanji were taught for reading and writing over six years of compulsory education. The average number of kanji students were able to write was approximately 500 (Gottlieb 1995, p. 144 & p. 159). In 1946, the Tōyō List, with a total of 1850 kanji, was introduced. It was deemed to have a close connection with script reform and efficiency of education (Gottlieb 1994, p. 1176). However, as a result of deliberations about how to teach these characters in school, the Tōyō Kanji Beppyo (Separate List of Characters for Interim Use) was announced, which is familiarly known as the kyōiku kanji (Characters for Education). It contains 881 characters of the Tōyō List to be taught for recognition and production over nine years of compulsory education (Gottlieb 1995, pp. 143-144). The question then arises of whether the students were able to recognise and produce the 881 kyōiku kanji at the time. In 1960, The National Language Research Institute surveyed regional primary schools. Results demonstrated that the average number of characters that students in their sixth year of primary school could write was 573 (Gottlieb 1995, p. 159). However, from this, it can be observed that students were not meeting the expectations of reading and writing of kanji (i.e., kyōiku kanji) determined by policymakers in the survey results from the early stages of script reform.
In 1981, the Tōyō List was altered to the Jōyō Kanjihyō (List of Characters for General Use), or jōyō kanji, which consists of the same characters, except for an additional 95, increasing the total to 1945 characters (Gottlieb 1995, p. 194). This reflected a change in the nature of the list, from a restriction to a guideline (Gottlieb 1994). The 881 kyōiku kanji, which remained the same, were to be taught over a shorter period of time—over the six years of primary school (Gottlieb 1995, p. 194). The kyōiku kanji was revised to the increased number of 1006 in 1989 (Mitamura & Mitamura 1997, p. ix). Today, the same 1006 kyōiku kanji designated to be taught over six years of primary school can be found. However, to ‘know’ kyōiku kanji suggests to both read and write, not by the end of primary school, but by middle school (i.e., compulsory education), as shown in MEXT’s guidelines (MEXT 2009). Therefore, another functional level of Japanese is knowing kyōiku kanji in terms of reading and writing. However, it is difficult to find evidence of these guidelines being met.
It is impossible to gauge the average number of kanji people can write by hand since the introduction of jōyō kanji due to the lack of data. This is an important distinction in the expected standard between kyōiku kanji and jōyō kanji. The MEXT’s guidelines indicate that students are to learn to read and write kyōiku kanji by the end of middle school, unlike jōyō kanji. Contradicting the first baseline of literacy previously mentioned, this is what shall be referred to as the second baseline of literacy (to know kyōiku kanji). Further, it shall be argued that there is yet another baseline of literacy set by MEXT.
The third baseline of literacy: Jōyō kanji
The third baseline of literacy is the expectation that students are to know the jōyō kanji. However, it has never been explicitly disclosed that to know means both to read and write. Rather, it is simply that to know means to be able to read jōyō kanji by the end of middle school (MEXT 2009). In addition to learning kyōiku kanji, the number of jōyō kanji that middle school students are required to learn to read are 300 to 400 characters for the first year, 350 to 450 characters during second year and all remaining jōyō kanji in the third year (MEXT 2009). Unlike year levels for primary school, in which specific kyōiku kanji are to be taught in each year, the kanji taught in middle are not indicated. It is not until senior high school that students are told to learn to write jōyō kanji over the three years. This does not align with actual practice; in general, middle school teachers of kokugo teach students both to read and write all jōyō kanji, or currently shin-jōyō kanji, throughout middle school (however, it was commented that due to classroom time constraints, in many cases students are instructed to study a number of these kanji outside of class, for example, as homework)2. In 2009, the government altered the jōyō kanji to the new jōyō kanji, but this did not change the standard students are to reach by the end of middle school. The kyōiku kanji remained the same and the number of the new jōyō kanji students are to know (i.e., be able to read) was changed from 1945 characters to 2136. Similar to the case for kyōiku kanji, evidence of students able to read the prescribed jōyō kanji is difficult to find.
More recently, the factor of technology further complicates the issue of literacy standards. In a survey by the Agency for Cultural Affairs (2001, cited in Gottlieb 2011), close to 90 percent of Japanese respondents perceived their ability to write in their own language as declining; almost 70 percent believed the same for their reading skills; and over 40 percent indicated their ability to write kanji had decreased due to using electronic media. While opinions vary whether the use of computers attributed to the decline in kanji proficiency shown in anecdotal evidence and surveys (Gottlieb 2011), Mino (2005, cited in Gottlieb 2011) observed students’ inability to read or write kanji properly without the assistance of a machine in classes at universities now that assignments no longer have to be hand written, as many institutions by the mid-1980s had replaced classes centred on teaching Japanese writing techniques with focus on the use of word processing.
About the Japanese-Language Proficiency Test
The JLPT was introduced in 1984. Only those of foreign citizenship were allowed to take it, aligning with the ideology of kokugo for Japanese people and nihongo for foreigners during the flourishing period of nihonjinron discourse. The JLPT is divided into different components: kanji, vocabulary, grammar, reading, and listening. The original JLPT comprised four levels; level 4 was the lowest and level 1 was the highest. In 2010, the JLPT changed to five levels, from N5 (the most basic level) to N1 (the most advanced level). Examinees of the JLPT are to choose the level they are to undertake upon application. Passing scores vary according to the level. The absence of any speaking or written component in the JLPT test is considered problematic (Nabei & Busch 1999). However, claims of progress, particularly with the new JLPT, shrouds its deficiencies.
Nabei and Busch (1999) highlight that the issue of the validity of the JLPT lies in the way the test’s content, or ‘syllabus’, was determined. They explained that this was done by conducting extensive surveys and frequency counts of vocabulary, kanji, and grammatical items found in Japanese-language textbooks and national curriculum standards created by the Ministry of Education for schooling in Japan. After, the items were allocated into what was perceived as progressively higher proficiency levels; that is, from the early stages of its existence, the JLPT’s content closely reflected the Japanese subject kokugo and the expectations set by MEXT.
A major factor in MEXT’s curriculum for kokugo, or course of study guidelines, is kanji. In a statistical analysis, the connection is strong. Using Pearson’s correlation coefficient (a statistical measurement for the relationship between two factors), it was demonstrated that the levels of the JLPT and school grades—or the kanji assigned to year levels in Japanese schools—are strongly correlated (Tamaoka, Kirsner et al. 2002; Tamaoka, Makioka et al. 2016). Even after the new JLPT was introduced, levels of the old JLPT correspond to the levels of the new version (Japan Foundation & Japan Educational Exchanges and Services 2009, p. 4). Therefore, it would make sense that the levels of the JLPT are comparable with levels of kokugo (as a school subject) despite the JLPT’s revision. The vocabulary Japanese people are expected to learn is contrasted to the JLPT levels and comments on the little research in recent years in this area are evident as in some references that were referred dated back to 1984, 1974 and 1951 (see Oshio et al. 2008 pp. 81-82). This factor is worth noting, but is not measurable in terms of course of study guidelines for kokugo, as only written scripts are used to distinguish levels.
JLPT-related books and journals demonstrate similar connections with kokugo’s guidelines and the JLPT’s levels. For example, the JLPT’s test content specifications for the old JLPT show that level 1 contains nearly all kanji from the jōyō kanji list except for 19 kanji that were deleted from the JLPT’s highest level. The reason given was that these kanji were based on the previous tōyō kanji list and are rarely used (Japan Foundation & Association of International Education 1994). Handa (2011) explained that the JLPT’s level 2 standard of 1000 kanji corresponds to the kyōiku kanji. Level 1 is based on the 1945 jōyō kanji. An earlier source from not long after the introduction of the JLPT stated that the 1000 kanji in level 2 is close to 996 characters (currently 1006 characters) of the kyōiku kanji (Takebe 1986, cited in Sakai 1993). Therefore, JLPT’s level 2 or N2 is comparable to that of the second baseline of literacy and level 1 or N1 corresponds to the third baseline of literacy.
Claims of the Japanese-Language Proficiency Test: Measuring communicative competence
Despite the JLPT remaining the same paper test consisting of the same components and lacking speaking and writing components, claims of how the new JLPT has improved are focused on communicative competence. This is evident in the tab under ‘Four Key Characteristics’ and is explained on the website in writing and in the visual diagram shown in Figure 1.
Figure 1: The Japanese-Language Proficiency Test’s Claims for Communicative Competence
Source: Image retrieved from https://www.jlpt.jp/e/about/points.html.
The incongruity of this diagram is that it does not follow the communicative models found in the literature, such as the model of communicative competence by Bachman and Palmer (1996). Its interpretation of performing tasks is through the same components as the old version of the test—vocabulary, grammar, reading, and listening. Unsurprisingly, Usami (2012) highlighted that, although the new JLPT claims to be more communicative, how this is actually achieved in terms of communicative competence is unclear, especially due to the lack of performing components. In the New Japanese-Language Proficiency Test Guidebook found on the JLPT Website, references to various scholarly academics, especially those related to communicative testing, are only found in the Japanese version (Japan Foundation & Japan Educational Exchanges and Services 2009, p. 40). Information about these references and other details are excluded from the executive summary in English and French. Another exclusion from the JLPT Website is the Webpage dedicated to research called Kanren ronbun, kiji (Related papers and articles) on the bottom of the tabs on the left-hand side, which only exists in Japanese (https://www.jlpt.jp/reference/research.html). This is found under the Shorui, shiryou annai (books and references) tab. The articles that are available are mostly newsletter-like and do not provide substantial evidence of the JLPT integrating into communicative models or the Common European Framework of Reference for Languages (CEFR). However, although it may not be immediately clear, there are claims that the new JLPT is also influenced by the CEFR.
Claims of the Japanese-Language Proficiency Test: Aligning with the Common European Framework of Reference for Languages
With the introduction of the new JLPT, the ‘Summary of Linguistic Competence Required for Each Level’ and the ‘JLPT Can-do Self-Evaluation List’ was also released, reflecting a global trend in following task-based assessment similar to that of the CEFR. Academics may consider this the case, as Bučar et al. (2015, p. 459) commented that the ‘new version of the JLPT proficiency levels was indirectly influenced by the concept of the CEFR.’ What is less clear is that it was intended that the JLPT levels were aimed to comply with CEFR. After retiring from the Japan Foundation, Kakazu (2014, p. 33) briefly asserted that the new JLPT would discover links with CEFR. Likewise, there are numerous articles on the JLPT website referencing the CEFR, including those in the Japanese version of the Japanese-Language Proficiency Guidebook (Japan Foundation & Japan Educational Exchanges and Services 2009). However, this also falls short of integration into the CEFR, similar to the claims for increased communicative competence. To be successfully integrated to CEFR (and communicative competence), the JLPT would need to test productive components, both speaking and writing, at various levels. However, the JLPT ultimately remains as a paper test, unchanging ever since it was first introduced.
Research in communicative competence and the CEFR are not the only areas in which Japanese testing is lacking. Watanabe and Koyanagi (2014) highlight that other related studies in washback in the field of Japanese testing are surprisingly limited. Further, they comment that there is a clear lack of empirical studies on validity in general. As previously mentioned, Nabei and Busch (1999) point out the concern the consider the JLPT as a proficiency test as it more closely resembles an achievement test; this achievement function can be related to what Messick (1989) calls the ‘social consequences’ component of validity (i.e., how the tests is being used). Messick (1989) defines validity as the following:
the degree to which empirical evidence and theoretical rationales support the adequacy and appropriateness of the inferences and actions based on test scores or other modes of assessment. (p. 13)
More recently, Hatasa and Watanabe (2017) commented that there are only 18 articles about empirical studies on Japanese testing research published since 1980 (of which 11 concern the JLPT) that demonstrate results of validity investigations. Similarly, it has been proposed that, although ‘testing and assessment have been a significant topic in L2 Japanese education, the number of studies that systematically evaluate testing or assessment instruments or processes themselves is still limited’ (Mori & Mori 2011, p. 475).
Since the introduction of the new JLPT, evaluation of the test is conducted by the Japan Foundation. However, test items and tasks are no longer shown and an annual report that was released prior to the new JLPT is no longer available (Watanabe & Koyanagi 2014). Test content specifications for the new JLPT are also no longer published; the reason given is that the goal is to communicate using the language rather than memorisation, as found on the JLPT’s ‘Frequently Asked Questions’ (FAQ) page (JLPT n.d.c). In short, the new JLPT has failed to integrate into more current models of testing—communicative competence and the CEFR—and a lack of transparency surrounding the test is evident. Further, in general, the validity of Japanese-language tests are limited.
Analysing the Japanese-Language Proficiency Test: Change in the examinees
With claims that the new JLPT fails to integrate communicative testing models and the CEFR, the question arises: what really has changed, if anything, concerning the JLPT? Rather than the test, other factors surrounding the test must be closely analysed. On the official JLPT Website, there is no information regarding exactly who is able to undertake the JLPT in terms of citizenship. On the FAQ page, the only comment on who can take the test is that the ‘JLPT is open to all non-native Japanese speakers’ (JLPT, n.d.c). However, there is information that provides more detail in the FAQ section of the Japanese-Language Proficiency Test Official Practice Workbook created by the Japanese Foundation and JEES:
I hold Japanese citizenship but my native language is not Japanese. Can I take the test?
The test is open to all non-native speakers of Japanese. It does not matter whether or not you hold Japanese citizenship. As the situation language is used varies on the person, whether or not that person’s native language is Japanese will be determined by applications made to the institution(s). Please consult the institution(s) if you are unsure of anything (Author’s translation) (Japan Foundation & Japan Educational Exchanges and Services 2012, p. 97).
This information is only written in Japanese (with furigana added to all kanji), therefore only those who are able to read Japanese can understand this statement. Unlike the original JLPT, it appears that the new version of the JLPT allows even Japanese citizens to undertake the test. What is more noteworthy is the information found on the ‘Advantages of JLPT’ page on the JLPT Website. On the second-last point of this Webpage, it is indicated that non-Japanese students are able to use the JLPT as accreditation for kokugo examinations in their final year of middle school. The following English statement demonstrates this: ‘the Japanese-language test is waved [sic] for examinees of foreign citizenship who pass JLPT N1 or N2’ (Author’s emphasis) (JLPT n.d.d).
One slight difference is excluded on the English version of the ‘Advantage of JLPT’ Webpage. The same Webpage in Japanese says ‘those with foreign citizenship, and so on, who have sat and passed N1 and N2 of the JLPT may use it to be exempt from kokugo exams’. (Author’s translation and emphasis), (JLPT n.d.e). Adding ‘and so on’ implies that this exemption does not necessarily mean that only those with foreign citizenship are eligible. This is also a common pattern of showing that such language policies are for foreigners only when one carefully examines information as found connected on the link to MEXT’s language policies on the ‘Advantage of JLPT’ Webpage—again, this information shown by MEXT is written in Japanese restricting readers to only those with Japanese proficiency.
On MEXT’s Webpage about the exemption examinations (for any year since 2011), it is worth briefly pointing out a statement under the first subheading, ‘According to the standards of School Education Act. Article 18, accredited middle school graduation examinations, for unavoidable reasons such as sickness, and so on…’ (Author’s translation and emphasis) (MEXT 2012). Reasons such as sickness are not mentioned on the JLPT or any other Website, questioning if justification for the policies is more the latter rather than the former.
Throughout the Webpage, the policies continue to emphasise being for foreigners only as the term ‘those of foreign citizens’ are found five times, without a single mention of Japanese citizens, before coming to the main section under the seventh heading ‘A part of exemption for subjects of examination, and so on’. This begins with the following statement:
In cases where a prescribed condition is satisfied, it is possible to apply for exemption, and so on, concerning part/s of the subjects for examination. For such measurements, there are all examinees that are targeted and examinees of foreign citizenships only, and so on, that are targeted… (Author’s translation and emphasis) (MEXT 2012)
It seems somewhat strange to have ‘all examinees that are targeted’ and ‘examinees of foreigner citizenship’ separated, however the former is found only in the very next statement, mentioning only that ‘anyone can apply’ for exemption for the English subject. Citizenship is clearly a subordinate factor, declaring that anyone with the following may be used as exemption for the English subject: any level from Level 3 up to Level 1 of The Eiken Test in Practical English Proficiency, Level 3, 2 or 1 of Eiken (English proficiency test) or any level of the United Nations Associations Test of English) (MEXT 2012).
After information about the English subject, the following section begins with declaring those who are applicable, again, commenting on only those with foreigner citizenship. It briefly states that the following is for the subject kokugo as follows:
(2) Those eligible who are examinees only of foreign citizenship, and so on. Restricted to those of foreign citizenship, and so on, who are examinees, it is possible to use both or one of the exemption options for kokugo in exceptional measures, which are to undertake examination papers with furigana added and by passing the ability examination. (Author’s translation and emphasis) (MEXT 2012)
The emphasise on foreign citizenship is quickly contradicted by the very next section, as found marked under the katakana character for ‘a’ or ア:
(ア) Examinees that are applicable
1) Those who do not have Japanese citizenship (either foreign citizenship or no citizenship).
2) Those who have Japanese citizenship and have lived abroad for at least two years.
3) Those who have been naturalised in Japan, and so on (Author's translation and emphasis) (MEXT 2012).
The second condition also states that, in addition to the requirement of living abroad for at least two years, the person has returned to Japan within a certain date—these dates have been set to be three years prior to the date the exemption examinations would be held. Put differently, this group is aimed for returnee children. The addition of ‘and so on’ at the end of the third condition makes allowances for the possibility of other cases to be eligible for the policies. This is the single section that refers to Japanese citizens also being applicable for the kokugo exemptions, reflecting the political tension behind the creation of the policies.
The next section reverts back to the distinctive pattern of referring to foreign citizens followed by ‘and so on’ marked under the katakana character for ‘i’ orイ:
(イ) Details on the exceptional measures
Undertaking examinations with furigana added examination papers
Examinees who are of foreign citizenship, and so on, are able to undertake examinations for all subjects with test papers that have furigana added for all kanji (excluding questions that ask readings for kanji)
2) Exemption for kokugo examination by passing the ability examination
Examinees of foreign citizenship, and so on, who have passed the ability examination in the following table may be exempt from the kokugo examination upon request. (Author’s translation and emphasis) (MEXT 2012).
For each point, an explanatory note is followed denoting ‘Necessary certificates’: ‘Necessary certificates from examinees of foreign citizenship, and so on (those applicable as mentioned above under (ア) number 2 or 3 are required. Those application as mentioned above under (ア) number 1 is not required’ (author’s translation and underlining) (MEXT 2012). It is quite ironic first to mention necessary certificates for foreigners, only to find that when one closely looks at (ア) number 2 or 3, neither would fall into such category.
Finally, a table follows, showing details that it is possible to use N1 or N2 of the JLPT (or level 1 or 2 of the original version) in lieu of the subject kokugo rather clearly. What is not quite clear are the conditions that even those with Japanese citizenship are eligible for exemptions for kokugo, with repeated referrals made to those of foreign citizenship throughout the Webpage. These language policies are the first to address Japan’s minority groups on a national level—something that has been long overlooked—by making adjustments for those applicable to graduate from compulsory education. With an estimate of 2,152,973 non-Japanese citizens living in Japan, as well as an estimated 12,000 Japanese returnees (exact number unknown), as pointed out by Okano and Tsuneyoshi (2011, p. 6 cited in Okano , 2017, p. 154), this is an issue that cannot simply be ignored.
MEXT emphasises that these language policies are aimed at foreign citizens. However, due to the increasing blurring of boundaries between kokugo and nihongo and the government’s reluctance to address minority groups in Japan, it is clear that these language policies were constructed to be as obscure as possible. This also contributes to the continuation of the nihongojinron discourse, despite changes accommodating minorities.
The question then arises: why did MEXT select the JLPT’s levels N2 or N1 (or level 2 or level 1 of the JLPT)? This ties back to the second baseline of functional literacy previously explained; the ‘ideal’ level of language proficiency set by MEXT that students are to know kyōiku kanji by the end of their compulsory education, which is also the kanji necessary to pass N2 (or level 2). To exceed this level would mean a more functional level of language proficiency; therefore, N1 (or level 1), which contains close to all jōyō kanji (or new jōyō kanji), or the third baseline of functional literacy, is included in the exemption.
The reasoning behind why MEXT created the option for undertaking all examinations, including kokugo, with furigana, connects back to their interpretation of literacy since postwar Japan; that is, the first baseline of functional literacy. In other words, MEXT considers that, at the very minimum, one is literate if they only have knowledge of hiragana. Since arguably high literacy rates have been around in Japan, this has always been the most vague level, as this very minimal level of literacy and its definition aligns with unrealistically high figures. These options in the newly derived language policies are no exception, as one must read carefully to discover these details. Unlike the JLPT, which demonstrates the marks required to pass each of its levels, it is unclear what score is required to pass the kokugo examinations for the third year of middle school.
This first baseline of functional literacy is what the government considers a functional level required for citizenship in Japan. Gottlieb (2011, p. 145) explains that:
Japan’s current language requirements for citizenship stipulate an ability to read and write Japanese equivalent to that gained by the second or third year of elementary school, a level of literacy clearly insufficient to allow full participation in public life.
In Japan, one learns hiragana during their first and second year of primary school. Therefore, a person would be deemed familiar with hiragana by the end of second year or the beginning of third year. Such a level would hardly comply with a realistic definition of literacy. For example, Taylor and Taylor (1995, p. 364 cited in Gottlieb, 2005, p. 92) describe the definition of functional literacy as the following:
People can read such everyday reading materials as newspapers and manuals, and also can fill in forms and write memos or simple letters. In addition, they have basic numeracy skills; and some can use a computer. They are likely to have finished at least middle school education.
Carroll (2001, p. 115) added that a reasonable definition of functional literacy in the Japanese context would be the ability to read a newspaper, which is a standard aimed for by the Ministry of Education by the end of compulsory education. However, the first baseline of functional literacy has been the bare minimum level that MEXT has upheld for decades. Realistic figures of Japanese literacy would date prior to the 1950s, as described earlier.
Although these language policies have been in place since 2011, a lack of transparency surrounding Japanese testing, along with vaguely set out language policies, has resulted in a grave lack of attention. However, one article that addresses these language policies is from Chapple (2014), who highlighted that these policies are examples that ‘reflect positive, and requisite, steps forward’, as furigana added to examinations demonstrates a shift away from traditionally stringent regulations and MEXT acknowledging kokugo and nihongo as synonymous for the first time. Citizens being allowed to undertake the JLPT and the test being allowed to be used as exemption for kokugo examinations—both considerable changes—are steps towards accommodation in language policies for minorities. However, this information has been done in a far from straightforward manner. Further, to acknowledge that JLPT’s N2 or level 2 is the equivalent of third year kokugo (and more significantly, furigana added to kokugo examinations) is far from what is required for those in need of language assistance. This raises the question of whether those who undertake the newly introduced options are linguistically prepared fully to function in Japanese society.
To add to the continuing nihonjinron discourse, there is a peculiar statement on the JEES Website describing the levels of the JLPT that is also connected to a link on the JLPT’s FAQ Webpage:
English statement: How many levels are there in the JLPT? How do these levels correspond to the Japanese school system?
There are 5 levels N1-N5 in the JLPT. Level N1 is the most difficult and level N5 is the easiest. As the JLPT Summary of Linguistic Competence Required for Each Level is different from the National Study Curriculum for native Japanese students, we cannot compare. This is because studying Japanese as one’s native language and as a foreign language requires different approach. (Japan Educational Exchanges and Services n.d.)
Japanese statement: How many levels are in the test? Approximately which year levels of Japanese schools align for each level of the test?
There are five levels, being N1 to N5, where N1 is the most difficult and N5 is the easiest. The certified standards of the test differ from the learning criteria for Japanese people (the Japanese language that Japanese people learn as a native language is different in learning content, order, and so on, to the Japanese language that foreigners learn as a foreign language), therefore the two cannot be compared. (Author’s translation and emphasis) (Japan Educational Exchanges and Services n.d.)
The Japanese version is conveyed in language that would feel unnatural to a native speaker of Japanese, as it is written in Easy Japanese, a concept that was originally intended for foreigners better to understand Japanese (see Burgess 2012). It is convenient for JEES to use Easy Japanese, as more complex words, such as ‘linguistic competence’ and ‘national study curriculum’ are excluded in the Japanese version, blurring the link between the new JLPT levels and year levels of Japanese schooling. This statement is typical of nihonjinron discourse, in which nihongo and kokugo are considered separate entities.
Despite the levels of the JLPT being comparable to Japanese schooling, the distorted justification for such a statement stems from the difference in the kanji allocated in levels of the JLPT. Interestingly, the difference between kokugo and nihongo education—or the levels between the JLPT and the levels of Japanese school (for the subject kokugo) and their relevance to Japanese-language testing—may be more recent than anticipated. In 1983, a report was released by the Agency for Cultural Affairs concerning Japanese-language proficiency for foreigners. The report indicated that 2000 kanji included the 1945 jōyō kanji plus 55 other kanji to comprise Dai-ichi suijun (level 1 standard). This was to be the standard of JLPT’s level 1. Here, Sakai (1993, pp. 111-112) states, is when the difference between kokugo education and nihongo education was born. This difference would continue to exist in the JLPT’s levels with never exactly the same jōyō kanji allocated in N1 or level 1 or kyōiku kanji allocated in N2 or level 2. However, the comparison would continue to be made among language educators, as previously shown.
This paper has demonstrated that the deficiencies in the JLPT—especially the new version—make it difficult to prove the progress of the test, particularly claims of communicative competence and alignment with the CEFR. The change in the JLPT’s new version that was introduced in 2010 has little to do with improvements in the test and is more concerned with the change in allowing Japanese citizens also to undertake it. This change coincided with the national introduction of language policies in the following year—a first of its kind—to address minority groups in Japan. Having explained the history behind kokugo versus nihongo, this represents the insider-outsider tenets becoming increasingly blurred. If anything, the JLPT remains to be a test comparable to kokugo, the school subject and its year levels set by MEXT.
While it is hard to determine the effectiveness of the newly emerged language policies, it is obvious that its utilisation is not widespread. This is unsurprising, given the lack of attention it has received and the far from straightforward fashion MEXT dispensed information concerning the language policies. The number of applicants was small as apparent in the most recent (and only available) figures—in 2018, the number of examinees who applied was 96, of which 72 passed and were able to graduate from middle school and allowed to continue onto senior high school (MEXT 2018).
Furthermore, the future of the progress in Japanese language testing instruments seem to be bleak. Not only have the operators of the JLPT reduced information on its procedures, such as no longer releasing test content specifications, but the staff number for its testing centre is notably low. At the Education Testing Service, which operates English tests such as TOEFL (Test of English as a Foreign Language) and TOEIC (Test of English for International Communication), approximately 1000 people are staffed in its main office; on the other hand, not even 100 people are on staff for Japanese testing (Kakazu 2014, p. 37). The combination of lack in transparency and professionalism has contributed to limited progress in Japanese testing. Therefore, much greater levels of transparency and efforts towards integration with modern testing models from testing-related organisations and support by government bodies are essential to make progress in this area.
Moreover, there may be additional difficulties created as a result. This includes difficulties to those non-native speakers of Japanese whose Japanese may not be at a level to function in Japan’s society even if they pass their examinations under accommodation of the new language policies. Similar accommodation or assistance is not available in senior high school, questioning whether or not applicants would be capable of keeping up with their studies at this level. Therefore, the changes involving the JLPT and language policies that are supposedly meant to accommodate minorities in Japan may be very much contradictory given the underlying political tension. It is argued that the JLPT itself has essentially not changed, but more importantly, utmost vigilant care has been taken as to maintain the divide between kokugo for Japanese people and nihongo for non-Japanese and thus the wider discourse of nihonjinron. In turn, this exacerbates the overriding issue of effectively assisting in the integration of minority groups into Japanese society.
In summary, the recent adjustments made in the JLPT and its newly emerged language policies reflect resistance to change in Japan’s identity by acknowledging minority groups, while simultaneously carefully crafting an image that portrays a linguistically and culturally homogenous nation.
Special thanks goes to Emeritus Professor Nanette Gottlieb, the University of Queensland, and Professor Ikuo Kawakami, Waseda University, for reading over the draft of this paper and providing comments. Their feedback and support is very much appreciated.
Bachman, LF & Palmer, AS 1996, Language testing in practice: Designing and developing useful language tests, vol 1, Oxford University Press, United Kingdom.
Befu, H 2001, ‘Hegemony of Homogeneity’, An Anthropological Analysis of Nihonjinron, Melbourne, Trans Pacific Press.
Brown, RA 1991. ‘Literacy assessments in polyscriptal societies: Chinese Character literacy in Korea and Japan’, Visible Language, vol. 25, no. 1, pp. 18.
Bučar, CS, Ryu, H, Škof, NM & Sangawa, KH 2015, ‘The CEFR and teaching Japanese as a foreign language’, Linguistica, vol. 54, no. 1, pp. 455-469.
Burgess, C 2012, ‘It’s better if they speak broken Japanese: Language as a pathway or an obstacle to citizenship in Japan?’ in N Gottlieb (ed.), Language and citizenship in Japan, Routledge, New York.
Carroll, T 2001, Language planning and language change in Japan, Curzon, Richmond.
Chapple, J 2014, ‘Japan’s immigration intimations and their neglected language policy requisites’, Asian and Pacific Migration Journal, vol. 23, no. 3, pp. 345-360.
Gottlieb, N 1994, 'Language and politics: The reversal of postwar script reform policy in Japan,' The Journal of Asian Studies, vol. 53, no. 4, pp. 1175-1198.
———1995, Kanji politics: Language policy and Japanese script, Routledge, United Kingdom.
———2005, Language and society in Japan, Cambridge University Press, United Kingdom.
———2008, ‘Japan: Language policy and planning in transition’, Current Issues in Language Planning, vol. 9, no. 1, pp. 1-68.
———2011, Language policy in Japan: The challenge of change, Cambridge University Press, United Kingdom.
Hadamitzky, W & Spahn, M 1996, Kanji & kana, Tuttle Publishing, United States.
Handa, A 2011, ‘漢字指導の留意点: 日本語教育の立場から [Points to remember about kanji instruction: The point of view from Japanese education]’, 全国大学国語教育学会発表要旨集 [The Japanese Teaching Society of Japan], no. 121, pp. 21-24.
Hatasa, Y & Watanabe, T 2017, ‘Japanese as a Second Language Assessment in Japan: Current Issues and Future Directions’, Language Assessment Quarterly, vol. 14, no. 3, pp. 192-212.
Heinrich, P 2011, ‘Difficulties of establishing heritage language education in Uchinaa’, in P Heinrich & G Galan (eds.), Language Life in Japan, Routledge, United Kingdom.
Japan Educational Exchanges and Services n.d, 2. 試験問題について [2. About the test] Retrieved from http://info.jees-jlpt.jp/faq/test_papers.html#, accessed 2 August 2018
Japan Foundation & Association of International Education 1994, 日本語能力試験 出題基準 [Japanese-Language Proficiency Test: Test Content Specifications], Bonjinsha Inc, Tokyo.
Japan Foundation & Japan Educational Exchanges and Services 2009, 新しい「日本語能力試験」ガイドブック[The New Japanese-Language Proficiency Guidebook], Japan Foundation & Japan Educational Exchanges and Services, http://www.jlpt.jp/e/reference/pdf/guidebook1.pdf, accessed 2 August 2018
Japan Foundation & Japan Educational Exchanges and Services 2012, 日本語能力試験公式問題集 N1. [Japanese-Language Proficiency Test Official Practice Workbook N1], Bonjinsha Inc, Tokyo.
JLPT (n.d.a), Homepage, http://www.jlpt.jp/e/index.html, accessed 29 July 2018
——— Objectives and history, http://www.jlpt.jp/e/about/purpose.html, accessed 29 July 2018
———FAQ, https://www.jlpt.jp/e/faq/index.html, accessed 29 July 2018
———Advantages of JLPT, https://www.jlpt.jp/e/about/merit.html, accessed 29 July 2018
———日本語能力試験のメリット[Advantages of JLPT], https://www.jlpt.jp/about/merit.html, accessed 29 November 2018
Kakazu, K 2014, ‘「日本語能力試験」をめぐる必然性 [Necessity of “Japanese-Language Proficiency Test”]’, 日本語教育 [Japanese Language Education], vol. 33, pp. 28-40.
Messick, S 1989, Measuring Second Language Performance, Macmillian, New York, pp. 13-103
Miller, RA 1982, Japan’s modern myth: The language and beyond, Weatherhill, New York and Tokyo
MEXT (Ministry of Education, Culture, Sports, Science and Technology) 2009, 学習指導要領「生きる力」[Learning guidelines and living strength], http://www.mext.go.jp/a_menu/shotou/new-cs/youryou/chu/koku.htm, accessed 9 August 2018
MEXT 2012, 中学校卒業程度認定試験 [Middle school graduation equivalent certification examinations], http://www.mext.go.jp/a_menu/shotou/sotugyo/1322849.htm, accessed 9 August 2018
———2018, 中学校卒業程度認定試験実施結果について [About the results of middle school graduation equivalent certification examinations], http://www.mext.go.jp/a_menu/shotou/sotugyo/1411519.htm, accessed 9 October 2019
Mitamura, JY & Mitamura, YK 1997, Kanji wo Benkyō Shimashō, Kodansha International, New York.
Mori, Y & Mori J 2011, ‘Review of recent research (2000-2010) on learning and instruction with specific reference to L2 Japanese’, Language Teaching, vol. 44, no. 4, pp. 447-484.
Nabei, T & Busch, M 1999, ‘The Japan Foundation and Association of International Education, Japan: Japanese-language proficiency test’, Applied Linguistics, vol. 20, no. 1, pp. 127-131.
Nakane, I, Otsuji, E & Armour, WS 2015, Languages and identities in a transitional Japan, in I Nakane, E Otsuji & WS Armour (eds.), Languages and identities in a transitional Japan: From internationalization to globalization, Routledge, New York.
Okano, KH 2012, ‘Languages and citizenship in education: Migrant languages in government schools’, in N Gottlieb (ed.), Language and citizenship in Japan, Routledge, New York.
Okano, K 2017, ‘Transnational migration and urban education in Japan’, in P Pink & G Noblit (eds.), Second international handbook of urban education, Springer International Publishing, Cham.
Oshio, K, Akimoto, M, Takeda, A, Abe, Y, Takanashi, M, Yanagisawa, Y, Iwamoto, R & Ishige, J 2008, ‘新しい日本語能力試験のための語彙表作成にむけて[Towards a new vocabulary list for the new Japanese-language proficiency test]’, 国際交流基金日本語教育紀要 [Japan Foundation Bulletin], vol. 4, pp. 71-86, http://www.jlpt.jp/reference/pdf/2008_010.pdf, accessed 29 July 2018
Sakai, J 1993, ‘漢字教育のシラバス (初級)―定着度をめぐる一考察― [Syllabus for the teaching of kanji at an intermediate level—a study of the degree of internationalization]’, Bulletin of Japanese Language Center for International Students, no. 19, pp. 109-139.
Seeley, C 1991, A history of writing in Japan, Brill, Leiden.
Shohamy, E 2006, Language policy. Hidden agendas and new approaches, Routledge, New York.
Sugimoto, Y 1999, ‘Making sense of Nihonjinron’, Thesis Eleven, vol. 57, pp. 81-96. doi: 10.1177/0725513699057000007
Tamaoka, K, Kirsner, K, Yanase, Y, Miyaoka, Y & Kawakami, M 2002, ‘A web-accessible database of characteristics of the 1945 basic Japanese kanji’, Behavior Research Methods, Instruments & Computers, vol. 34, no. 2, pp. 260-275.
Tamaoka, K, Makioka, S, Sanders, S & Verdonschot, RG 2016, ‘www. kanjidatabase.com: A new interactive online database for psychological and linguistic research on Japanese kanji and their compound words’, Psychological Research, vol. 81, no. 3, pp. 1-13.
Taylor, I & Taylor, M 2014, Writing and literacy in Chinese, Korean and Japanese: Studies in written language and literacy 14, 2nd edn, John Benjamins, Philadelphia. http://dx.doi.org/10.1075/swll.14
Tollefson, JW 1991, Planning Language, Planning Inequality, Longman, London.
Unger, JM 1996, Literacy and script reform in occupation Japan: Reading between the lines, Oxford University Press, New York.
Usami, Y 2012, ‘What is the “communicative language competence required to perform tasks” that is to be measured in the New Japanese-Language Proficiency Test?’ Japanese Studies Journal, vol. 29, no. 1, pp. 1-15.
Watanabe, N 2015, ‘Contextualizing current activities of the national language division in Japan’, 総合政策研究 [Journal of Policy Studies], no. 48, pp.93-104.
Watanabe, Y & Koyanagi, K 2014, ‘Assessing Japanese’, in AJ Kunnan (ed.), The companion to language assessment, John Wiley & Sons, Inc, United States. doi: 10.1002/9781118411360.wbcla048
Yamashiro, JH 2013, ‘The social construction of race and minorities in Japan’, Sociology Compass, vol. 7, no. 2, pp. 147-161.
Yoshino, K 2005, Cultural nationalism in contemporary Japan: a sociological enquiry, London, Routledge.
 In 2001, the Ministry of Education was renamed the Ministry of Education, Culture, Sports, Science and Technology.
 Interview with a teacher of the subject kokugo at Eimei Junior High School, Chino city, Nagano Prefecture, 3 February 2019
Article copyright Roxanne Lizelle Niveri and Sol Rojas-Lizana.