1 Introduction to the Present Volume
1.1 The Japonic Languages
The Japonic languages comprise Japanese and Ryukyuan and their respective local dialects (Figures 1.1 and 1.2).
Under the name of ‘Japanese’ are subsumed all local varieties spoken on Japan’s mainland (Mainland dialects) and a language spoken in Hachijōjima, which lies to the south of Japan’s mainland. Ryukyuan languages are spoken in an area that was once an independent country (Ryukyuan kingdom, 1429–1869), from the Amami islands to Yonaguni. There is no mutual intelligibility between Japanese and Ryukyuan, between Mainland dialects and Hachijō, or between Northern and Southern Ryukyuan languages. Even within Northern and Southern Ryukyuan, there is no mutual intelligibility among major island varieties. So, it is a recent shared recognition among Japanese linguists that Japonic comprises several distinct languages, Mainland language (which is ‘Japanese’ in a narrower sense), Hachijō, Amami, Okinawan, Miyako and Yaeyama.
Based on various lexical, phonological, and morphosyntactic variations, several major dialectal groupings (isoglosses) have been suggested (Tojo 1966, Kindaichi 1964, Hirayama 1968, inter alia). The foundational classification is Tōjō’s work, to which revisions, both major and minor, have been made subsequently by a number of researchers. According to Tōjō’s classification and Hirayama’s (1968) revision, the Japonic languages divide into three major dialectal groups: the Mainland dialects, the Hachijōjima dialect and Ryukyuan. The Mainland dialects further subdivide into (a) Eastern Japanese (E.Jpn) (Hokkaidō, Tōhoku, Kantō, Tōkai-Tōsan), (b) Western Japanese (W.Jpn) (Hokuriku, Kinki, Chūgoku, Shikoku), and (c) Kyūshū Japanese (K.Jpn). Ryukyuan subdivides into (a’) Amami, (b’) Okinawa, and (c’) Miyako, Yaeyama and Yonaguni.
A genetic subgrouping of the Japonic languages yields a different grouping of languages, according to which the Japonic family first divides into Japanese and Ryukyuan. Japanese further subdivides into Mainland dialects and the Hachijōjima dialect. Ryukyuan subdivides into Northern Ryukyuan (N.Ryu) and Southern Ryukyuan (S.Ryu). Northern Ryukyuan subdivides into Amami and Okinawan, while Southern Ryukyuan subdivides into Miyako and Macro-Yaeyama, which further subdivides into Nuclear Yaeyama and Yonaguni (Pellard 2015). (Hattori 1976) suggests a different hypothesis where Ryukyuan and Kyūshū dialects form a subgroup, which branched off from proto Kyūshū-Ryukyuan, a hypothesis which has recently been taken seriously by several scholars (see Igarashi 2021 for discussion).
1.2 The Focus of the Present Volume
Our focus in the present volume is not on Standard Japanese (SJ; the de facto standard language based on Tokyo Japanese), which is perhaps one of the best known and most oft-cited languages in typological literature. Rather, our exclusive focus is on the local varieties of Japonic, four from Mainland dialects (Nambu, Izumo, Yanagawa and Shiiba) and five from Ryukyuan (Tokunoshima, Iheya, Kin, Aragusuku and Shiraho) (see Figures 1.1 and 1.2). Local varieties of Japonic have hitherto been little known to the readers outside Japan, and have thus largely been ignored in typological studies. For example, the World Atlas of Language Structures (WALS, Haspelmath et al. 2005), which is the world’s largest typological database, does not include data on a single dialect of Japanese. This by no means indicates that the Japonic Family is typologically homogenous or that it suffices to take up SJ when making typological claims under the name of ‘Japanese’. On the contrary, just as SJ is a typologically interesting language and is worth a separate book, the various local varieties of Japonic are also worth a detailed typological characterization in their own right. In fact, a detailed look at the local varieties of Japonic often sheds new light on various typological claims which, as mentioned, have been made without reference to these languages. For example, as discussed in § 2.2.3 below, marked-nominative case alignment (Handschuh 2014), which is often regarded as a highly unusual alignment type cross-linguistically or even as an ‘unexpected’ type, is robustly observed in a number of Ryukyuan languages, casting strong doubts on the existing typological and theoretical assumption pertaining to case alignment which theoretically rules out such a system.
The purpose of the present volume is thus to serve as a useful and informative introduction to the typologically heterogeneous language family, the Japonic languages, by providing a collection of grammatical sketches of diverse Japonic languages both from Mainland dialects and Ryukyuan. Each chapter covers a wide range of descriptive topics that will aid the readers to have a broad picture of the phonological and morphosyntactic organization of each language (see the next section for the structure of each chapter). Our prospective readers are linguists based outside Japan: typologists, descriptive linguists working on neighboring languages, historical linguists and general linguists of any theoretical persuasion who are not expected to be familiar with Japanese linguistics. Since the target readers are not expected to be familiar with Japanese linguistics and its terminological tradition, the present volume carefully avoids making too much use of the local linguistic terminology and tries to adhere to terms and concepts which are widely recognized in general linguistics, especially linguistic typology.
Table 1.1
Descriptive format for each chapter
|
Topics |
Subtopics to be addressed |
|
|---|---|---|
|
Section 1 |
The language and its speakers |
Geographic, genealogical and sociolinguistic information, etc. |
|
Section 2 |
Phonology |
Inventory of phonemes, syllable structure and phonotactics, word-level prosody, intonation, etc. |
|
Section 3 |
Descriptive units |
Word classes, segmentation (word, affix, clitic, etc.), grammatical relations, etc. |
|
Section 4 |
Nouns |
Pronouns, lexical nouns, numerals, their internal structure, etc. |
|
Section 5 |
Verbs |
Inflectional morphology, stem-internal structure, special grammatical verbs like the copula, etc. |
|
Section 6 |
Adjectival expressions |
Word-class status, inflected vs. non-inflected adjectives, etc. |
|
Section 7 |
Class-changing derivation |
Nominalization, verbalization, adjectivalization, etc. |
|
Section 8 |
Demonstratives and interrogatives |
Words for deixis and anaphora; interrogative words, indefinite words, etc. |
|
Section 9 |
Argument phrase |
Internal structure of NP, case markers, headless structures, etc. |
|
Section 10 |
The simple sentence |
Sentence types, alignment, possession, valency-changing, TAM, information structure, etc. |
|
Section 11 |
The complex sentence |
Coordination, subordination, clause-chaining, etc. |
|
Appendix |
Sample text |
Fully-glossed spontaneous speech collected, transcribed and annotated by the author(s) |
1.3 The Structure of the Present Volume
The present volume consists of ten chapters. The present chapter has set the goal of the volume and will give a typological overview of the Japonic Family in the following sections with frequent references to the data and description in the subsequent chapters in the volume. Chapters 2 to 6 are grammatical sketches of Ryukyuan languages, and Chapters 7 to 10 are grammatical sketches of Mainland dialects. Genetic and geographical coverage have been taken into consideration in selecting which languages were taken up in the present volume. The five chapters on Ryukyuan languages cover both Northern and Southern Ryukyuan varieties and their major subgroups (Amami, Okinawan, Miyako and Yaeyama). The four chapters on Mainland dialects cover all three major dialectal areas, i.e. Eastern, Western and Kyūshū dialects.
Each chapter follows roughly the same descriptive format (Table 1.1), with minor adjustments where necessary depending on the language being described. The shared descriptive format enables readers to quickly search for typological features they are interested in and also to easily compare different languages with respect to specific typological topics such as syllable structure, prosody, word-class assignment, verbal inflectional categories, interrogative-sentence formation, etc. Note that each chapter includes a fully glossed sample text. It is intended as a sample of the way the language is actually used in natural discourse and as supplemental data which demonstrates the usage of morphemes, constructions, etc., described in each chapter.
2 A Typological Overview
This section aims to give a typological overview of the Japonic languages, focusing on both phonology and morphosyntax. It does not go into the phonological and grammatical detail which specialists of Japanese linguistics would expect to see. Rather, the present section serves to give the readers a basic understanding of the phonology and grammar of Japonic and pays attention to several typological highlights.
2.1 Phonology
Phonology is one major area where Japonic languages show conspicuous internal variation, although Japonic languages do share a number of basic typological features, which I briefly overview here before discussing dialectal variation below. First, the open light syllable CV is the most basic building block of the basic vocabulary. Second, word-medial consonant clusters are limited to, or at least include, geminates or partial geminates (e.g. homorganic nasal + C). Third, if a word-final coda is permitted, it must be, or at least include, the nasal /n/ (typically pronounced as [n] or [ŋ]). Fourth, there is no stress language among the Japonic languages, and word-level prosody of the Japonic languages is broadly characterized under the rubric of pitch-accent system (see § 2.1.4 for more detail).
2.1.1 Vowels
In SJ and most other Japonic languages, the length distinction between short and long vowels is robust, but this is not the case in some Mainland dialects and Ryukyuan. Nambu (see Nakagawa, this volume), for example, has no length distinction for vowels (although it has a singleton-geminate contrast, i.e. a contrast for consonants). In a few Southern Ryukyuan languages such as Shiraho (Urabe, this volume), the length contrast for vowels does seem to exist in some words (e.g., /turu/ ‘bird’ vs. /tuuruu/ ‘lamp’) but seems to be absent for others (e.g., /pitu/ [pitu] or [pituː]).
SJ and most other Mainland dialects have a five-vowel system with /a/, /e/, /o/, /i/ and /u/, although in some dialects neutralization may occur, especially after a coronal obstruent where neutralization of the high vowels is very common. For example, /i/ and /u/ is neutralized after coronal obstruents in Izumo (Western Japanese, Hirako, this volume), as in //das-u// (put.out-npst) and //das-i// (put.out-inf) which are both realized as [dasɨ]. See also Nakagawa (this volume) for a detailed description of the neturalization of /i/ and /u/ in Nambu. In addition to the five cardinal vowels, cross-linguistically less common vowels such as [ɛ], [æ], [œ], [ɜ], etc., may arise from diphthongs (hence occurring as long in many dialects), which may be due to a historical change or may be derived by a synchronic rule. For example, Tōhoku dialects such as Nambu (Nakagawa, this volume) typically has /ɛ/, which was formerly a diphthong such as /ai/, /oi/, /ae/, etc., as in /kɛna/ ‘arm’ (cf. a cognate of which, kaina, is found in other Japonic languages such as Miyako Ryukyuan). In Kisogawa (Eastern Japanese, Hirako et al. 2019), the vowels [æː] and [œː] are derived from underlying /ai/ and /oi/ respectively, as in //taka-i// (high-npst) ‘high’ → [takæː], //hoso-i// (thin-npst) ‘thin’ → [hosœː], etc.
Ryukyuan languages show more conspicuous internal variation regarding their vowel systems. Basically, /e/ and /o/ occur as long, as proto-Ryukyuan short *e and *o changed to /i/ and /u/ in Ryukyuan languages. Yonaguni (Southern Ryukyuan) has the smallest inventory of vowels, /a/, /i/ and /u/. Amami (Northern Ryukyuan), especially the north Amami languages, characteristically have central vowels /ɨ/ and /ə/.1 Southern Ryukyuan languages generally have a high central vowel which may carry a friction noise to different degrees depending on the dialect, as in Miyako [ps̩tu] and Ishigaki [pɨtu] ‘person’. The sound in question is typologically known as a fricative vowel (Ladefoged and Maddieson 1996, Shimoji 2006). In some languages such as Miyako (except for Ōgami), the phonetic fricative vowel has a dual function phonologically, making it difficult to analyze it as a phonemic vowel or consonant (see Shimoji 2006 for a brief summary of various analyses of the fricative vowel in Miyako). For example, in Aragusuku (see Wang, this volume), the fricative vowel /ž/ may occur as a syllable margin (/žža/ [zːa] ‘father’ (CCV), /paž/ [paz] ‘fly’ (CVC), etc.) or, with a restriction on the onset, may additionally occur as a nucleus (/pžtu/ [ps̩tu] ‘person’ (CC̩.CV)). Wang analyzes it as a syllabic consonant.
2.1.2 Consonants
SJ and most other Japonic languages have a two-way distinction between voiceless and voiced consonants for obstruents, but some have a three-way distinction between voiceless, voiced and prenasalized (e.g. /t/ [t] vs. /d/ [d] vs. /nd/ [nd]), as is commonly observed in the Tōhoku dialects (Eastern Japan; see Nakagawa, this volume), or between voiceless, voiced and laryngealized for stops (e.g. /t/ [t] vs. /d/ [d] vs. /ˀt/ [ˀt]), as in Ryukyuan, especially Northern Ryukyuan languages (see Katō, this volume).2 Ōgami (Pellard 2009; Miyako, Southern Ryukyuan) lacks a voicing contrast altogether for obstruents, where all obstruents are voiceless (/p, t, k, f, s/). Sonorants do not have a voicing contrast (always voiced phonetically), except that a language like Ikema (Hayashi 2013; Miyako, Southern Ryukyuan) has a contrast between /n/ and /n̥/ (/nna/ [nːa] ‘conch’ vs. /n̥na/ [n̥ːa] ‘rope’). Shiraho (Urabe, this volume) has a similar contrast between /n/ and /n̥/.
Japonic languages have a phoneme which reflects proto-Japonic *p, which later lenited to /h/ via ɸ word-initially and to /w/ (or zero) medially in most Japonic languages. Some Mainland dialects such as Tsugaru (Eastern Japanese, Aomori) and Shiiba (Kyūshū, see Shimoji and Hirosawa, this volume) have ɸ which reflects *p, as in Tsugaru [ɸebi] ‘snake’ (cf. SJ hebi), Shiiba /hwyaa/ [ʏ̥aː] ‘fly’ (cf. SJ hae), etc. A number of Ryukyuan, especially in Southern Ryukyuan languages (see Wang, this volume, and Urabe, this volume) have /p/, as in /paa/ ‘leaf’ (cf. SJ ha), /paz/ ‘fly’ (cf. SJ hae), etc. The debuccalization process *p > ɸ > h caused a number of synchronic peculiarities of the phoneme in question in the present systems of many Japonic languages, in such a way that the labial feature still remains in some parts of the current language system.3 One such example is Rendaku (Sequential Voicing), where a voiceless obstruent (e.g. /t/) becomes its voiced counterpart (e.g. /d/) if it occurs as the initial onset of a non-initial root in a compound structure, as in SJ /hon-dana/ ‘bookshelf’ (//hon// ‘book’ + //tana// ‘shelf’). If the obstruent in question is /h/ underlyingly, then it alternates with /b/, as in SJ /hon-bako/ ‘bookcase’ (//hon// ‘book’ + //hako// ‘case’).
2.1.3 Syllable and Mora
SJ has a cross-linguistically common canonical syllable structure (C)(G)V(C) where V may be a long vowel or a diphthong, as in /ka/ ‘mosquito’, /kya.ku/ ‘guest’, /kai.ko/ ‘silkworm’, etc. The same canonical syllable structure is observed across most Mainland dialects, although some dialects show conspicuous divergence from this, as in Shiiba (Shimoji and Hirosawa, this volume) where there is a double glide sequence /wy/, which is phonetically realized as a non-syllabic (i.e. glide) version of the rounded front vowel [ʏ], as in /wyaata/ [ʏaːta] ‘boiled’ or /hwyaa/ [ʏ̥aː] ‘fly’.
Ryukyuan languages often have a more complex syllable structure with an onset cluster CC and/or with a special pre-onset slot which is filled only by a laryngeal ˀ. In Miyako, for example, there is a wide range of consonants which may form an onset geminate CC with the first C being a moraic onset, as in /ffa/ [fːa] ‘child’, /ssan/ [sːaŋ] ‘not know’, /vva/ [ʋːa] ‘2sg’, /mma/ [mːa] ‘mother’, etc. (see Wang, this volume). It is a controversial issue whether onset geminates can be moraic (cf. Hayes 1989), but the Ryukyuan data suggests that they may be moraic in some languages (and other languages such as Trukese, Topinzi 2008). Miyako further has a rich array of syllabic consonants, and some of them may even carry an onset, as in /pztu/ [ps̩tu] ‘person’ (CC̩.CV), etc.
The mora plays a central role in most Japonic languages in such a way that the length contrast between long and short vowels and between singleton and geminate consonants is phonemic. Prosodic rules also make much more reference to the mora than the syllable. Such significant roles of the mora over the syllable are not clear in Tōhoku (see Nakagawa, this volume) and southern Kyūshū, which are collectively known as ‘syllabeme’ languages in Japanese linguistics as opposed to ‘moraic’ languages. Western Japanese dialects typically have a bimoraic minimality constraint (BMC) whereby a word in isolation must have at least two morae. This constraint is also observed in Ryukyuan languages in general. Languages may differ with respect to what domain the BMC applies to: a word or to a word plus clitic (Shimoji 2010), a noun or a noun plus a noun modifier within an NP (Matsuoka 2021). Another factor which plays a role in the application of the BMC is the type of clitic which attaches to the host noun: Matsuoka (2021) demonstrates that certain types of clitic (e.g. case particle) tend to be integrated into the domain of the BMC while others (e.g. copula, sentence-final particle, etc.) do not.
2.1.4 Prosody
As mentioned in the introduction to the present section, word-level prosody of Japonic has traditionally been characterized as a pitch-accent system (see Uwano 2012 for an overview), in which an accented word contains one (and only one) prominent mora (or syllable), which causes a certain pitch event (abrupt fall or rise). In SJ, for example, the accented trimoraic word /tamágo/ ‘egg’ (where the accented mora is marked) is pronounced as [tàmáŋò] (%LHL), with the initial %L being a phrase-initial boundary tone and the final L being a result of the abrupt fall in pitch after the accented mora /ma/. In some Japonic languages, the accented mora causes an abrupt rise rather than fall, either of the mora itself or after it. In Nambu (Eastern Japanese; see Nakagawa, this volume), for example, /manágu/ ‘eye’ is pronounced as [mànáɡɯ̀] (LHL%) and /kendó/ ‘road’ is pronounced as [kèǹdô] (LLHL%, where the final H and L% docks onto the final mora /o/ and are realized a rise-fall contour). Notice that the mirror image situation of SJ is found with regard to boundary-tone assignment, whereby the final lowering L% lowers the pitch of the final mora.
In this way, pitch accent is characterized by its culminative nature, i.e. the existence of one prominent mora/syllable within the word domain, a feature which makes the pitch-accent system look like a stress system. However, the pitch accent system of Japonic differs from a stress system in that in the former, a word may crucially lack a prominent mora, i.e. a word may be ‘unaccented’ in Japanese linguistics terms, as in SJ /sakana/ [sàkáná] (%LHH) ‘fish’ and in Nambu /nazugi/ [nànzɯ̀ɡí] (LLH) ‘forehead’.4 In a stress system, an unstressed lexical word is usually ruled out, i.e. there must be one prominent syllable/mora within a word. In prosodic typology, a stress system and a tone system constitute two distinct linguistic types and a pitch-accent system is regarded as a hybrid of the two, not a distinct ‘type’ (Hyman 2006). As a typological hybrid, then, the various prosodic systems which are collectively called pitch-accent systems of Japonic languages may show a considerable internal variation, and the variation can be captured in terms of what kind of stress-like features and tone-like features are present in the system.
Some Japonic languages exhibit a feature which is characteristic of tone languages, i.e. the paradigmatic choice between specific tonal melodies for each lexeme, e.g. between ‘Level’ and ‘Rising’ in Kyoto Japanese (Western Japanese), or between ‘Level’ and ‘Falling’ in Kagoshima Japanese (Kyūshū Japanese). Tone as is found in these systems is called ‘word tone’ (Hayata 1998), as the domain of the paradigmatic tonal contrast is the word rather than the syllable/mora. However, these systems still show a stress-like character in that a certain mora/syllable is specified as prominent, i.e. accented, although the nature and function of accent differs between the two languages. In Kyoto, a lexeme is specified for tone and prominence (accent), with both having a contrastive function. In Kagoshima, a word is specified for tone only, and a post-lexical and non-contrastive rule determines which syllable/mora is accented to realize the lexically specified tone (Falling or Level). For example, in Kagoshima Japanese, the word /onago/ ‘woman’ has a Falling tone (/onago/[F]) while the word /otoko/ ‘man’ has a Level tone (/otoko/[Lv]). A post-lexical rule then determines where the fall occurs (penultimate syllable), with the output for /onágo/[F] becoming [ònágò]. In the case of the level tone, the final syllable is assigned H by a post-lexical rule. The Kagoshima type is especially common in Kyūshū and Ryukyuan languages (see Kato, Tamamoto, Carlino and Urabe, this volume).5
A few Japonic languages have a characteristic typically associated with a stress language, i.e. a foot-based metrical structure. This is true, for example, the Aoya dialect of Tottori (Western Japanese, Matsumori 2012), the Maisaka dialect of Shizuoka (Eastern Japanese, Poppe 2016), the Nakijin dialect of Northern Ryukyuan (Lawrence 1990), the Nagahama dialect of Irabu (Miyako, Southern Ryukyuan), etc. However, such languages are much less well documented than tone-like pitch accent languages as noted above.6
2.2 Morphosyntax
This section discusses the morphosyntactic features of Japonic. From a broad range of topics pertaining to morphosyntax, this section focuses on the number category, adjectival expressions and alignment systems, since these in particular exhibit a surprising amount of internal diversity, are well documented in our volume, and are worth the special attention of theoretical linguists and typologists. There are of course many more typologically interesting topics which are not covered in this section simply due to limitations of space, and the reader is referred to § 3 of the present chapter to see other topics covered in the present volume.
Before going into detail, let us briefly note morphosyntactic features which are shared by all Japonic languages and therefore show no variation. Japonic languages are strictly verb-final, with a modifier-head order and dependent marking. Morphological organization is overwhelmingly suffixing and agglutinating, although verbs and adjectives may show a fusional character due to their complex inflectional morphology, especially when compared with nouns. Besides affixation, compounding and reduplication are common word-formation strategies.
2.2.1 Number
As in SJ, most Japonic languages have a dichotomic number system in which the singular and the plural are distinguished, while some Ryukyuan languages, especially Amami (Northern Ryukyuan), have a trichotomic system where the singular, the plural and the dual are distinguished for personal pronouns. In Tokunoshima (Amami, Northern Ryukyuan), for example, first- and second-person reference distinguishes between the singular (e.g. first-person wan), the plural (waakja/wakkja) and the dual (wanten/wattari), while third-person reference is made by using demonstrative pronouns which show a dichotomic number contrast between the singular and the plural (see Kato, this volume).
In SJ and all other Japonic languages, number marking is sensitive to the lexical properties (especially animacy) of the noun to which number marking applies in two major ways: obligatoriliness and semantic interpretation of plural marking.
Regarding obligatoriliness, the number distinction between singular (zero) and non-singular is obligatory only for pronouns and a limited set of lexical nouns, or address nouns (Shimoji 2022), which can be used as terms of address. Here, the lack of non-singular marking indicates singular denotation, as in SJ watasi-∅ ‘I’ vs. watasi-tati ‘we’. All chapters of the present volume report that the number distinction is obligatory for pronouns. Many report that address nouns also show the same property, as in Tokunoshima (Northern Ryukyuan, Kato, this volume), Shiiba (Kyūshū, Shimoji and Hirosawa, this volume) and Yanagawa (Kyūshū, Matsuoka, this volume). In contrast, for other classes of noun, a lack of non-singular marking may be a mere underspecification of number, as in SJ kodomo ‘child/children’, whereas an explicit marking of non-singular is always interpreted as non-singular, as in SJ kodomo-tati ‘children/*child’. Inanimate nouns are usually not subject to number marking at all, although there are languages like Shiraho (Urabe, this volume) which allow plural marking of both animate and inanimate nouns (e.g. hanako-nda ‘Hanako and others’, sara-nda ‘dishes’, etc.). Typically, if an inanimate noun takes plural marking, the effect is ‘quasi-pluralization’, whereby a certain set of referents are denoted in the way the focal referent represents it as an exemplar (see Niinaga 2020 for a typological survey of this phenomenon as observed in Japonic). In Shiiba (Kyūshū Japanese, see Shimoji and Hirosawa, this volume), for example, a plural-marked inanimate noun functions to denote an exemplar of a certain unspecified set, as in sumoo-domo (Sumo.wrestling-pl) ‘Sumo wrestling and suchlike’.
Regarding the semantic interpretation of plural marking, animacy plays a significant role in the interpretation of associativity, a distinction between additive and associative plural. The additive plural has to do with a homogeneous pluralized set, as in SJ gakusee-tati (student-pl) ‘students’, while the associative plural has to do with a heterogeneous set of entities which consists of a focal referent and his/her associates, as in SJ taroo-tati (Taro-pl) ‘Taro and his associates’. Whereas SJ and most other Mainland dialects do not have a formal distinction between additive and associative plurals and the animacy of the noun serves as an indirect indicator of associativity (i.e. the higher the animacy of the noun, the more likely it is to denote associative plural), a number of Ryukyuan languages have differential plural-marking patterns which serve to disambiguate the distinction between associative vs. additive interpretations. For example, in Kin (Okinawan, Northern Ryukyuan; see Tamamoto, this volume) and Iheya (Okinawan, Northern Ryukyuan; see Carlino, this volume), there are two plural markers, (a) -ta (Kin)/-taa (Iheya) as a general plural marker which can be used for a wide range of nouns and is not specified for associativity, and (b) -ntja (Kin)/-nčaa (Iheya), which is attached to a much more restricted set of human nouns (which pertain to certain age, sex, kinship-role groups). (b) is used to denote the additive plural only (e.g. Kin ikiga-ntja ‘males’). To the best of my knowledge, if a Ryukyuan language has a strategy to disambiguate the distinction between associative and additive plural, the system always has a special additive plural marker in addition to a general plural marker, and not a special associative plural marker.7 Nakagawa (this volume) reports a special plural marker -ho in Nambu (Tōhoku, Eastern Japanese), which denotes a specific group (family, village, organization, etc.) to which the focal referent being marked with this suffix belongs. She lists examples like ora-ho (1-gr) ‘we (our family)’, which is in contrast in form and function to ora-ndo (1-pl) ‘we’. Whereas the denotation of the first-person plural is simply the speaker + other referents, the denotation of the first-person ‘group’ further specifies the way referents are united as a group.
The distinction in clusivity for the first person non-singular, i.e. exclusive and inclusive, is very common in Ryukyuan, while it is totally absent in Mainland dialects.8 The typology of clusivity (Filimonova 2005) has revealed that there are two major types of languages which distinguish between the exclusive and the inclusive: (a) languages which have the two distinct forms and (b) languages which have a dedicated inclusive form but lack a dedicated exclusive form. Type (b) is usually understood as a system where ‘I’ and ‘we:excl’ are neutralized (with both denoting a referent/referents without the addressee). In our data, both types are documented, with Shiraho (Yaeyama, Southern Ryukyuan; see Urabe, this volume) exemplifying (a) and Aragusuku (Miyako, Southern Ryukyuan; see Wang, this volume) exemplifying (b). However, the Aragusuku system diverges from the canonical type (b): in Aragusuku, the first-person plural category contrasts between the general (which may be used for either exclusive or inclusive reference) and the inclusive.9
2.2.2 Adjectival Expressions
The typology of adjectival expressions in diverse languages (Dixon 1977, 2004, Wetzer 1996, etc.) has revealed that adjectival expressions may show considerable cross-linguistic variation with respect to their word-class assignment. Japanese (i.e. SJ) is understood as a cross-linguistically uncommon ‘split-adjectival’ system in terms of Wetzer’s (1996) typology of adjectival expressions, where two major lexical classes of adjectives co-exist within a single language system, i-adjectives (inflected/verbal adjectives) and na-adjectives (non-inflected/nominal adjectives). For example, taka- ‘high’ is an i-adjective root while kiree is a na-adjective root, and which root belongs to which class is lexically determined. i-adjectives are like verbs as they inflect for tense (e.g. taka[-i] ‘high (present)’ vs. taka[-katta] ‘high (past)’). However, they differ from verbs in the kind of inflectional suffixes they take: in SJ, for example, the non-past affirmative inflection of the verb is -ru, while that of the i-adjective is -i, etc. na-adjectives (e.g. kiree ‘beautiful’) are like nouns in that they do not inflect and require a copular verb for inflection (kiree⸗da[-tta] ‘beautiful (past)’). However, the na-adjectives differ from nouns since they cannot occur as arguments.
The above-mentioned character of the split-adjectival system found in SJ largely holds true for most, if not all, Mainland dialects (see the four grammatical sketches of Mainland dialects in this volume), with some minor dialectal variation mentioned below. For example, some Mainland dialects, especially Kyūshū dialects, have a small class of roots which may be coded verbally (as a verbal adjective) or nominally (as a nominal adjective). In Yanagawa (Kyūshū Japanese, see Matsuoka, this volume), a root like kiree- ‘beautiful’ may be coded as a verbal adjective with tense inflection, e.g. kiree-katta (beautiful-pst), or as a nominal adjective with a copular verb following it, e.g. kiree⸗yat-ta (beautiful⸗cop-pst).
In Ryukyuan languages, a lexical class distinction such as that between i- and na- adjectives in SJ is largely irrelevant, with all adjectival roots being treated homogeneously. The single class of adjective roots is usually treated like verb roots, in the sense that they inflect for tense, mood, etc. Unlike SJ and many other Mainland dialects, the inflectional suffixes used with verbal adjectives are identical to those used with (existential) verbs, making the verbal adjective one step closer to the verb class. Tokunoshima (Amami, Northern Ryukyuan; see Kato, this volume) and Shiraho (Yaeyama, Southern Ryukyuan; see Urabe, this volume) treat verbal adjectives in their respective languages as a subclass of verbs. In some Ryukyuan languages, especially in Miyako, a given adjectival root may be coded either nominally or verbally, demonstrating a cross-linguistically rare ‘switch-adjectival’ system in Wetzer’s (1996) typology of adjectival expressions. The difference between switch- and split-adjectival systems is that in the former, one and the same adjective root may be coded either nominally or verbally. For example, in Aragusuku (Miyako, Southern Ryukyuan; see Wang, this chapter) a given adjectival root may carry a verbalizer suffix -kar- which further carries a verbal inflectional suffix, as in taka-ka-taa (high-vlz-pst) ‘was high’, while the same root may be suffixed with the morpheme -munu, as in taka-munu (high-munu) ‘(sthg is) high’. The munu form exhibits a nominal characteristic in that it may be used in construction with copula. The morpheme munu is undergoing a grammaticalization process whereby the nominal compound root munu ‘thing’ is being recategorized as a nominal-adjective derivational suffix -munu (see Shimoji 2009).
A switch-adjectival system is also reported in the Kin and Iheya dialects of Okinawan (Northern Ryukyuan; see Tamamoto and Carlino, this volume). As mentioned above, Kyūshū dialects like Yanagawa show a similar fluctuating coding pattern for a restricted set of adjectival roots, and the difference between Ryukyuan switch-adjectival systems and that of Yanagawa is that in Yanagawa, the set of adjectival roots which exhibit a switch-adjectival character is much more restricted and is considered a historical shift from nominal adjectival roots to verbal adjectival roots.
2.2.3 Alignment
One prominent feature of the alignment patterns found in Japonic is that all alignment patterns except ergative-absolutive are attested, i.e. nominative-accusative, neutral, split intransitive and tripartite, demonstrating the typological diversity of the Japonic languages. Nominative-accusative alignment is found across Mainland dialects and Ryukyuan. Neutral alignment is found in Hateruma (Yaeyama, Southern Ryukyuan; Aso 2020), which has no core case marker and word order is an important way to identify S, A and P. Neutral alignment is also partially integrated in the alignment system of a number of Japonic languages. For example, Nambu (Tōhoku, Eastern Japanese; see Nakagawa, this volume) displays an animacy-driven split of alignment whereby animate nouns show a nominative-accusative pattern where A/S is zero coded and P is marked with accusative ⸗ba, as illustrated in (1a), while inanimate nouns are typically zero-marked and may yield a neutral pattern, as in (1b).
(1)
a.
taro
{wa/omɛ/hanako/tomodati/inu}⸗{ba/??∅}
mi-tera.
Taro
{1sg/2sg/Hanako/friend/dog}⸗{acc/∅}
see-prog
Taro is looking at {me/you/Hanako/(his) friend/a dog}.
b.
taro
sodo⸗{ba/∅}
mi-tera.
Taro
outside⸗{acc/∅}
see-prog
Taro is looking outside.
This kind of animacy-driven overt P marking is widely known as Differential Object Marking (Bossong 1985), and is common in Eastern Japanese.
Yonaguni (Macro-Yaeyama, Southern Ryukyuan; Shimoji 2016) exhibits a split-intransitive pattern. In Yonaguni, A is always marked with ⸗nga, which is cognate with SJ ga, while P is always zero-marked. With regard to S, a certain lexically determined set of intransitive verbs such as MOVE, GO, WALK, etc. (unergative, motion verbs) always require S to be marked with ⸗nga, while another set which designates change-of-state events like BREAK, FALL, CRACK, BEND, etc. (unaccusative, change-of-state verbs) always require S to be zero-marked. Other intransitive verbs may align either way. Thus, Yonaguni has a lexical split of S into Sa and Sp, which are coded like A and P respectively, demonstrating a Split-S as opposed to a Fluid-S system (Dixon 1994).
Tripartite alignment is very rare in Japonic, but is attested in a very small number of Kyūshū dialects. It occurs as a result of an animacy-driven split of alignment. In Shiiba (Shimoji and Hirosawa, this volume), for example, alignment splits into nominative-accusative, split-intranstive and tripartite depending on the animacy of the S/A/P. As summarized in Table 1.2, down to the lower end of Animacy Hierarchy, animal nouns take different marking for A (⸗ga), S (⸗ga or ⸗no) and P (ba/⸗oba/⸗o). Note also that in the middle of Animacy Hierarchy, human nouns show a split-intransitive alignment pattern where ⸗ga covers A and Sa while ⸗no covers Sp only. Unlike a typical pattern found in the world’s languages where there are distinct case forms for S, A and P, the distinction between A and S in Shiiba involves the difference in terms of whether the argument in question may take either ⸗ga (A) or ⸗ga plus ⸗no (S). Thus, there is no dedicated case marker for A.
Table 1.2
Alignment patterns of Shiiba
|
Pronoun |
Human |
Animal |
Inanimate |
|
|---|---|---|---|---|
|
A |
G |
G |
G |
N |
|
Sa |
G/N |
|||
|
Sp |
G/N |
|||
|
P |
B/OB/O |
B/OB/O |
B/OB/O |
B/OB/O |
Another feature which is worth typologists’ attention is the internal diversity of nominative-accusative systems exhibited in Japonic languages, where all possible nominative-accusative alignment patterns are attested: marked nominative-accusative where both are marked, as observed in SJ, marked accusative where only P is marked (see (1a)), and marked nominative where only S/A are marked, as illustrated in (2) from Iheya (Okinawan, Northern Ryukyuan; see Carlino, this volume).
(2)
a.
taruu⸗ga
ačč-oo-ta-n.
Taro⸗nom
walk-prog-pst-ind
Taro (S) is walking. (Intransitive)
b.
taruu⸗ga
hasi
wata-ee
nz-a-n.
Taro⸗nom
bridge
cross-seq
go-pst-ind
Taro (A) crossed the bridge (P). (Transitive)
Cross-linguistically, marked nominative is known to be extremely rare (Handschuh 2014: 1). However, Iheya and two other Northern Ryukyuan languages included in the present volume display a marked-nominative alignment system: Kin (Okinawan, Northern Ryukyuan) and Tokunoshima (Amami, Northern Ryukyuan). All Okinawan dialects and a few Amami dialects such as Okinoerabu (Yokoyama 2017) exhibit the same marked-nominative pattern where S/A are marked by the nominative ⸗ga or ⸗nu and P is always left unmarked.
Marked nominative is unusual or even unexpected if we assume that core-case marking exists to distinguish between the two core arguments of a transitive clause, i.e. A and P (Comrie 1978, Hoop and Malchukov 2008, etc.). According to this widely-held assumption, there should be no functional motivation for S being overtly case-marked. It is therefore predicted that if there is an unmarked case for any of S, A, and P, then it must be used (at least) for S, a prediction known as Greenberg’s Universal 38 (Greenberg 1963). The prediction is then made that if A or P is left unmarked in a system, S must also be unmarked. The marked-accusative pattern (where only P is marked) and the marked-nominative-accusative pattern (where S, A, and P are all marked) do not contradict this prediction, whereas marked-nominative alignment clearly goes against this prediction.
There is one thing which may be a key to understand the underpinnings of marked-nominativity in Japonic, which pertains to a well-known typological characteristic of this language group, i.e. its topic-prominent feature. In all languages of Japonic, whether they have marked-nominative alignment or not, nominative-case marking and topic marking ⸗wa are in a paradigmatic relationship.10 That is, nominative marking not only marks a grammatical relation (S/A) but an information-structural status of non-topic (Shimoji 2018). Nominative indicates that the argument so marked is not a topic. S/A are relatively more likely to be interpreted as a topic, so nominative marking may serve as a useful alert of S/A being non-topic against hearer’s expectation. Crucially, especially in those languages which are claimed to be marked nominative (Okinawan), certain kinds of S strongly tend to be left unmarked (i.e. without nominative marking) even if they are not a topic, and Shimoji (2018) has shown that such S is inanimate and unaccusative S, which is low in topicality and thus are not expected to be a topic of a sentence, making the non-topic marking (i.e. nominative marking) dispensable.
The ostensible ‘mystery’ of why S (and A) is marked may thus become a well-motivated coding strategy if we carefully examine marked nominativity exhibited in Japonic languages especially in terms of information structure, but much is left for further investigation. Focusing on Okinawan and other marked nominative languages of Japonic will be a promising field of research for the typology of marked nominativity.
3 A Brief Summary of the Subsequent Chapters
Chapter 2 is a grammatical sketch of Tokunoshima (Amami, Northern Ryukyuan), focusing on the Isen dialect. It has a number of features which are substantially different from other Amami dialects or even from other areal varieties in the island in terms of vocabulary, phonology, and morphosyntax, such as an ongoing merger of front and mid vowels, residual intervocalic /r/ and marked-nominative system.
Chapter 3 is a grammatical sketch of Iheya (Okinawan, Northern Ryukyuan). Iheya is an endangered language with a speaker population of about 400, mostly in their 60s and older. The chapter is one of the first comprehensive descriptions of an Okinawan variety besides Shuri, the prestige dialect spoken in what used to be the capital of the Ryukyu Kingdom. Covering a wide range of descriptive topics, it uncovers features unreported in the Ryukyuan languages until now, such as the use of the reflexive pronoun duu as a second-person pronoun. It also includes a description of intonation, which is still under-described in many Japonic varieties, focusing on interrogatives, which have as an unusual feature a non-rising pitch when forms with an interrogative meaning are used.
Chapter 4 is a grammatical sketch of Kin (Okinawan, Northern Ryukyuan). Kin is in many ways distinct from well-known varieties of the Okinawan language such as Shuri. Topics which have been little discussed in the literature and will be covered extensively in this chapter include: tone assignment (in which both notions of mora and syllable are relevant), adjectival morphology (in which an alleged adjectival “word” is decomposed into two components, i.e., a grammatically independent stem and an inflectional clitic), and declarative uses of the non-finite connective form of verbs (whose semantics are related to the continuative aspect and mirativity.)
Chapter 5 provides a grammatical sketch of Aragusuku (Miyako, Southern Ryukyuan). In addition to describing basic phonological and morphosyntactic characteristics of Aragusuku, the chapter addresses several typological issues particular to Aragusuku or the Miyako language in general. For examples, the fricative vowel (e.g. [ksks] (CC̩.CC̩) ‘listen’), differential reflexive marking whereby different reflexive forms are used depending on person, plurality and case, the double-subject construction which is sensitive to the kind of possessive relationship between the two subject NPs (e.g. karjaa miinudu kagimunu. ‘As for him, (his) eyes are beautiful.’, etc.)
Chapter 6 gives an overview of the grammar of Shiraho (Yaeyama, Southern Ryukyuan). Shiraho village was formed as a result of massive migration from Hateruma Island during the early eighteenth century. With this peculiar demographic history and its later language contact with the neighboring villages speaking other Yaeyaman dialects, Shiraho occupies a unique position within Yaeyama in terms of phonology, morphology and vocabulary. Notable features are as follows: (1) a sonorant voicing contrast (/n/ vs. /n̥/), (2) a plural suffix which can attach to non-animate nouns unlike in other Japonic languages, (3) an aspectual distinction expressed using pitch, (4) two types of past tense (simple past form and converbal form).
Chapter 7 is a grammatical sketch of Nambu (Tōhoku, Eastern Japanese). It has a trichotomic contrast between voiceless, voiced and prenasalized obstruents (e.g. /akeru/ [akeɾu] ‘openable’ vs. /ageru/ ‘open’ vs. /angeru/ ‘raise’). Both A/S and P are frequently zero-coded, and P may be overtly marked with the accusative ⸗ba if it is animate, a phenomenon cross-linguistically known as Differential Object Marking. In addition to causativization and passivization, which are common in all Japonic langauges, there is a third productive valency-decreasing strategy, anticausativization, whereby a transitive verb is intransitivized to construe an event as a spontaneous event with no agency or control.
Chapter 8 is a grammatical sketch of Izumo (Umpaku, Western Japanese). Despite the fact that Izumo is geographically a dialect of Western Japanese, it shares a number of features with Eastern Japanese dialects, such as the existence of the copula ⸗da, leading to a long-standing controversy among linguists with regard to how Izumo is situated in the history of Japanese. The chapter is the first detailed grammatical description of this dialect written in English, covering a broad range of phonological and grammatical topics based on the author’s own fieldwork. It addresses typologically interesting features such as a (morpho-)phonemic alternation involving the liquid /r/ (e.g. /ku-ru/ [kwaː] come-npst), zero nominalization, as in jai-ta⸗o kuu-ta (bake-pst⸗acc eat-pst) ‘(I) ate the baked (one)’, differential nominative/genitive-case marking which is sensitive to the animacy of the NP, etc.
Chapter 9 gives a grammatical sketch of Yanagawa (Kyūshū Japanese). Yanagawa is spoken by the older generation (in their seventies or older) and is in imminent danger of extinction. The chapter covers a range of descriptive topics and discusses a number of typologically interesting features such as symmetrical differential subject marking (two nominative particles, ⸗ga and ⸗no, which are used according to the animacy of the S/A argument), minimal-word constraints sensitive to various syntactic conditions (such as the presence or absence of cliticization), morphophonological rules which refer to bimoraic foot structure, etc.
Chapter 10 is a grammatical sketch of Shiiba (Kyūshū Japanese), with a special focus on the Omae dialect, which is one of the most endangered dialects of Shiiba with a local population of approximately 400. The chapter includes descriptions of typologically uncommon features of this language such as the existence of the complex glide onset /wy/ (which may even be combined with the glottal /h/, as in /hwyaa/ [ʏ̥aː] ‘fly’), Differential Object Marking which is sensitive to the animacy of both A and P, whereby P must be overtly marked if the animacy of P outranks that of A, an Experiencer Construction where the stimulus argument is non-canonically marked with dative, potentiality marking which distinguishes between ‘situation-driven’ potentiality and ‘ability-driven’ potentiality, etc.
Articulatory and acoustic analysis of the mid central vowel shows phonetic variation. See Kato, this volume, for his detailed description of this and the other central vowel in Tokunoshima.
Nakagawa (this volume) notes that it is difficult to find minimal triplets to justify the three-way distinction in Nambu. It may be alternatively analyzed as a two-way system where the voiceless vs. voiced contrast is now reanalyzed as geminate vs. singleton, as in /tt/ [t] vs. /t/ [d], especially given the fact that in this language, the voiceless sound tends to be pronounced with a longer duration, as in [odotːo] ‘younger brother’, which is interpreted by Nakagawa as /odoto/ but may be /ototto/ by the present author’s alternative analysis.
In SJ, several scholars analyze what is usually analyzed as /h/ as /p/, which is realized as [h] by rule, an analysis which captures the labial feature still present in SJ. See, for example, McCawley (1968).
The final H for /gi/ in the Nambu data must be due to a predictable post-lexical rule which signals a boundary tone.
Miyakonojō Japanese has an extremely small range of paradigmatic choice of tone, i.e. Level tone only (Kibe 2010: 25), which is realized as a tonal contour with H on the word-final syllable (/onago/ [ònàgó] ‘woman’, /otoko/ [òtòkó] ‘man’, etc.). Yanagawa (Matsuoka, this volume) and Shiiba (Shimoji and Hirosawa, this volume) further lack a post-lexical fixed accent rule, resulting in a prosodic system where there is no consistent pitch pattern. Diachronically speaking, it is one possible scenario that the lexically contrastive word-tone system changed to the Miyakonojō type, which further changed to the Yanagawa-Shiiba type where the tonal information has been lost altogether.
Bimoraic feet are reported to play a significant role in the pitch accent of a number of Japonic languages as mentioned above. In the domain of morphophonology, the importance of bimoraic feet as a morphological template has been extensively documented, both in SJ (e.g. Poser 1990, Kubozono 1993) and in various Japonic languages, including Yanagawa (Western Japanese, Kyūshū; Matsuoka, this volume), where clipping of unfooted segments occurs in the derivation of specific verb forms. Also, the bimoraic minimality constraint as observed in Western Japanese and Ryukyuan (§ 2.1.3) can be analyzed as a constraint which requires a word to contain at least one foot.
Yoron (Amami, Northern Ryukyuan) is another such example (Kibe et al. 2019). In Yoron, reaccentuation occurs to denote the additive plural meaning as opposed to associative plural meaning, as in [ʔatɕa-taa (father-PL) ‘father and his associates’ → ʔatɕa[taa (father-PL) ‘fathers’, where [ indicates an accented mora after which abrupt rise occurs. The reaccentuation induces an additive plural interpretation where an associative interpretation is usually expected, as in proper names and elder kinship terms. It thus reflects the functional markedness of the additive plural interpretation for address nouns.
It is often pointed out that SJ does have the distinction in clusivity with the humble first person plural watasi-domo, which usually excludes the addressee. However, one can say watasi-domo mo mairimasyoo ‘Let’s us (incl) go’. Let us imagine a situation where there are a master and two of his servants with an equal rank and the master left them. The sentence above is uttered by the servants to each other, i.e. with inclusive reference, meaning they (the two servants) should follow him (the master). The upshot is that the humble -domo has a pragmatic effect of excluding the addressee by its very nature of humble marking: the speech participants are divided into the honored person and the speaker who gives such honor. Usually, the honored person is the addressee, leading to exclusive reference. But in a situation like the one I set above where the person to be honored happens to be absent in the speech situation but where the use of humble marking is still the norm, the humble watasi-domo ‘we’ can easily be inclusive, demonstrating that it is semantically not an exclusive pronoun but a mere humble marker.
A few Okinawan dialects of Ryukyuan such as Jana have a distinction similar to clusivity, in which three referential categories are distinguished: (a) exclusive ‘we’, (b) inclusive ‘we’ which is in contrast with a third person (as in ‘They’re going to eat noodle, and what are we:incl going to eat then?’) and (c) inclusive ‘we’ without contrast with a third person. The common feature of (a) and (b) is ‘contrast’ (with the addressee in the case of (a), or with the non-locutor in the case of (b)). In Jana, the same form (wattaa, which is cognate with an exclusive pronoun in neighboring dialects) is used for (a) and (b) while a different form (agamii, which is cognate with an inclusive pronoun in neighboring dialects) is used for (c). A clusivity system would distinguish between (a) and (b)/(c), but the Jana system crosscut the clusivity distinction, distinguishing between (a)(b) and (c). See Shimoji (2021) for details of this system and its typological characteristics.
In most Ryukyuan languages, the nominative may cooccur with the topic, as in Miyako mma⸗ga⸗a (mother⸗nom⸗top), but in this case the topic is always a contrastive topic.
Bibliography
Aso, Reiko (2020) Minami Ryūkyū Yaeyamago Hateruma Hōgen no Bumpō [A grammar of Hateruma, a Yaeyama dialect of Southern Ryukyuan]. PhD thesis, Tokyo University of Foreign Studies.
Bossong, Georg (1985) Differentielle Objektmarkierung in den Neuiranischen Sprache. Tubingen: Gunter Narr Verlag.
Comrie, Bernard (1978) Ergativity. In Lehman, Winfred P. (ed.) Syntactic Typology. Austin: University of Texas Press, 329–394.
Dixon, R.M.W. (1977) Where have all the adjectives gone? Studies in Language, 1: 19–80.
Dixon, R.M.W. (1994) Ergativity. Cambridge: Cambridge University Press.
Dixon, R.M.W. (2004) Adjectives classes in typological perspective. In Dixon, R.M.W. and Alexandra Y. Aikhenvald (eds.) The Adjective Classes. Oxford: Oxford University Press, 1–49.
Filimonova, Elena (2005) Clusivity: Typology And Case Studies of Their Inclusive-exclusive Distinction. Amsterdam: John Benjamin.
Greenberg, Joseph (1963) Some universals of grammar with particular reference to the order of meaningful elements. In Greenberg, Joseph (ed.) Universals of human language. Massachusetts: MIT Press.
Handschuh, Corinna (2014) A typology of marked-S languages. Berlin: Language Science Press.
Haspelmath, Martin, Matthew S. Dryer, David Gil, and Bernard Comrie (2005) The World Atlas of Language Structures. Oxford: Oxford University Press.
Hattori, Shiro (1976) Ryūkyū hōgen to hondo hōgen [The Ryukyu dialects and the mainland dialects]. In Kinenka, Ifa Fuyu Seitan Hyakunen (ed.) Okinawa Gaku no Reimei: Ifa Fuyu Seitan Hyakunen Kinenshi [The dawning of Okinawan studies: A volume commemorating the hundredth anniversary of the birth of Ifa Fuyu]. Okinawa: Okinawa Bunka Kyōkai.
Hayashi, Yuka (2013) Ryūkyū Ikema Hōgen no Bumpō [A grammar of Ikema Ryukyuan]. PhD thesis, Kyoto University.
Hayata, Teruhiro (1998) Akusento to tōn: inritsu no ippanron ni mukete [Accent and tone: towards a general theory of prosody]. Ajia Afurika Bunpō Kenkyū [Asian and African Linguistics], 26: 1–22.
Hayes, Bruce (1989) Compensatory lengthening in moraic phonology. Linguistic Inquiry, 220: 253–306.
Hirako, Tatsuya, Ai Kubozono, and Kyoji Yamaguchi (2019) Kisogawa Hōgen Bunpō Gaisetsu [A sketch grammar of Kisogawa dialect]. In Aoi, Hayato and Nobuko Kibe (eds.) Kisogawa Hōgen Chōsa Hōkokusho [A research report of Kisogawa dialect]. Tokyo: National Institute for Japanese Language and Linguistics, 7–70.
Hirayama, Teruo (1968) Nihon no Hōgen [The Japanese dialects]. Tokyo: Kōdansha Gendai Shinsho.
de Hoop, Helen and Andrej L. Malchukov (2008) Case-marking strategies. Linguistic Inquiry, 39(4): 565–587.
Hyman, Larry (2006) Word-prosodic typology. Phonology, 23(2): 225–257.
Igarashi, Yosuke (2021) Bunkigakuteki shuhō ni motozuita nichiryū shogo no keitō bunrui no kokoromi [A cladistic approach to phylogenetic classification of Japanese and Ryukyuan]. In Hayashi, Yuka, Tomohide Kinuhata, and Nobuko Kibe (eds.) Fīrudo to Bunken kara Miru Nichiryū Shogo no Keitō to Rekishi [Genealogy and history of Japanese and Ryukyuan from the perspectives of philology and fieldwork]. Tokyo: Kaitakusha, 17–51.
Kibe, Nobuko (2010) Hōgen akusento no tanjō [The formation of various accent systems in Japanese]. NINJAL Project Review, 2: 23–35.
Kibe, Nobuko, Hajime Oshima, and Masahiro Yamada (2019) Plural Forms in Yoron-Ryukyuan and Address Nouns in Ryukyuan Languages. Japanese/Korean Linguistics, 25: 1–11.
Kindaichi, Haruhiko (1964) Watashi no hōgen kukaku [My version of dialectal classification]. In Kai, Nihon Hōgen Kenkyū (ed.) Nihon no Hōgen Kukaku [The Classification of Japanese Dialects]. Tokyo: Tokyodo, 71–94.
Kubozono, Haruo (1993) The organization of Japanese prosody. Tokyo: Kurosio.
Ladefoged, Peter and Ian Maddieson (1996) The Sounds of the World’s Languages. London: Blackwell.
Lawrence, Wayne (1990) Nakijin phonology: feet and extrametricality in a Japanese dialect. PhD dissertation, University of Tsukuba.
Matsumori, Akiko (2012) Nishi Nihon ni okeru nobori kaku no hōgen: Tottori Ken Aoya Chō to sono shūhen chīki ni okeru akusento taikei [A rising kernel dialect and the accentual systems surrounding Aoya in Tottori Prefecture]. NINJAL Research Papers, 3: 19–37.
Matsuoka, Aoi (2021) Miyazaki Ken Shība Son Omae hōgen ni okeru saishōgo seiyaku: shūshoku yōso no umu, kōzoku suru yōso no shurui ni chakumoku shite [Word minimality constraint in the Omae dialect of Shiiba, Miyazaki Japanese: with a focus on the presence/absence of the modifier of a noun and the type of the subsequent element of a noun]. In Nihon Gengo Gakkai Dai 162 Kai Taikai Happyō Genkō Shū [Proceedings of the 162nd Conference of Linguistic Society of Japan]. Language Society of Japan, 5–11.
McCawley, James D. (1968) The Phonological Component of a Grammar of Japanese. Berlin/New York: Mouton de Gruyter.
Niinaga, Yuto (2020) Kita Ryūkyū Amami Ōshima Yuwan hōgen no meishi daimeishi fukusūkei no kinō to sono tsūgengoteki ichizuke [Non-number values of nominal and pronominal plural forms in Yuwan in crosslinguistic perspective]. Gengo Kenkyu, 157: 71–112.
Pellard, Thomas (2009) Ogami: Éléments de description d’un parler du sud des Ryukyus. PhD thesis, École des hautes études en sciences sociales.
Pellard, Thomas (2015) The Linguistic Archaeology of the Ryukyu Islands. In Patrick Heinrich, Shinsho Miyara and Michinori Shimoji (eds.) Handbook of the Ryukyuan Languages. Berlin/Boston: Mouton de Gruyter, 12–37.
Poppe, Clemens (2016) Iambic feet in Japanese: evidence from the Maisala dialect. Gengo Kenkyu, 150: 117–135.
Poser, William (1990) Evidence for foot structure in Japanese. Language, 66(1): 78–105.
Shimoji, Michinori (2006) Syllable structure of Irabu Ryukyuan. Shigen, 2: 21–40.
Shimoji, Michinori (2009) The adjective class in Irabu Ryukyuan. Nihongo no Kenkyuu [Studies in the Japanese Language], 5(3): 33–50.
Shimoji, Michinori (2010) Ryukyuan languages: an introduction. In Shimoji, Michinori and Thomas Pellard (eds.) An introduction to Ryukyuan languages. Tokyo: ILCAA, 1–11.
Shimoji, Michinori (2016) Minami Ryūkyū Yonagunigo no kaku hairetsu ni tsuite [Alignment in Yonaguni Ryukyuan]. In Ryūkyū Shogo to Kodai Nihongo [Ryukyuan and Classical Japanese]. Tokyo: Kurosio, 173–207.
Shimoji, Michinori (2018) Dialects. In Hasegawa, Yoko (ed.) Cambridge Handbook of Japanese Linguistics. Cambridge: Cambridge University Press, 87–113.
Shimoji, Michinori (2021) Kita Ryūkyū Okinawago Nakijin Jana Hōgen ni okeru jokatsusei [Clusivity in the Jana dialect of Nakijin Okinawan, Northern Ryukyuan]. Nihon Gengo Gakkai Dai 163 Kai Taikai Happyō Genkō Shū [Proceedings of the 163rd Conference of Linguistic Society of Japan], 178–184.
Shimoji, Michinori (2022) Number in Japonic Family. In Acquaviva, Paolo and Michael Daniel (eds.) Number in the world’s languages: a comparative handbook. Berlin/New York: Mouton de Gruyter, 505–528.
Tojo, Misao (1966) Kokugo no hōgen kukaku [The classification of the dialects of the national language]. Tokyo: Tokyodo.
Topinzi, Nina (2008) On the existence of moraic onsets. Natural Language and Linguistic Theory, 26: 147–184.
Uwano, Zendo (2012) Three types of accent kernels in Japanese. Lingua, 122: 1415–1440.
Wetzer, Harrie (1996) The typology of adjectival predication. Berlin/New York: Mouton de Gruyter.
Yokoyama, Akiko (2017) Ryūkyū Okinoerabujima Kunigami Hōgen no Bunpō [A Grammar of the Kunigami Dialect of Okinoerabu Ryukyuan]. PhD thesis, Hitotsubashi University.