1.1 Geography and Affiliation

View Full Size

Figure 8.1

Nita area

• Download Figure
• Download figure as PowerPoint slide

The Izumo dialect is spoken in the Izumo area, the east part of Shimane Prefecture, including Izumo City, Unnan City, Matsue City, Yasugi City, and Okuizumo Town. Shimane Prefecture is situated in the western part of the main island of Japan, so the Izumo dialect has been classified as a Western Japanese dialect since Tōjō (1953). However, it shares several 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.

According to Hiroto (1950), the Izumo dialect can be classified into three subdivisions: the Northeastern dialects, the Northwestern dialects, and the Southern dialects. These differences are based on their phonological, lexical, and grammatical features (Table 8.1). The present chapter focuses on the Nita dialect, one of the Southern dialects (Figure 8.1). The Nita dialect is spoken in the former Nita Town area of the present Okuizumo Town, located on the prefectural border of Shimane and Hiroshima prefectures.

1.2 Sociolinguistic Overview

The population of the former Nita area of Okuizumo Town is about 6,500 (as of 2020), but the number of people who can speak the traditional Nita dialect is much fewer. Fluent speakers of Nita are mostly over seventy years old, and many are in their eighties or nineties.

While able to understand the traditional dialect, the younger generations normally choose to speak Standard Japanese or use dialectal forms familiar to the whole region of Izumo, which is similar to the dialect spoken in Izumo City and distant from the traditional Nita dialect.

1.3 Previous Works

One of the earliest descriptive studies on the Izumo dialects is Katō (1935), a short description of the morpho-syntax of one dialect in Northeastern Izumo. Hiroto (1950) is the landmark study on Izumo, and it contains some descriptions of the phonology and morphology of the dialects spoken in Shimane and Tottori prefectures. Since then, no work has produced a comprehensive synchronic description of this dialect, although there have been some studies on individual phenomena, such as Hirako (2016) on case marking and Hirako and Tomosada (2018) on the morphology.

2.1 Phoneme Inventory

Nita has five vowels (/i[i̞ ~ i ~ ɨ], e[e ~ e̝], a, o[o ~ o̝], u[ɯ]/), thirteen consonants (/p, b; t, d; k, g; c[ts ~ tɕ], z[dz ~ dʑ]; s[s ~ ɕ], h[h ~ ɸ], m, n[n ~ m ~ ŋ ~ ɴ], r[ɾ ~ ɭ ]/) and two glides (/w, y[j]/). Phonetic long vowels are analyzed as a sequence of vowel phonemes: [aː] is /aa/, [oː] is /oo/, and so on. Closed vowels can be devoiced after a voiceless consonant, and non-closed vowels can also be devoiced between voiceless consonants (e.g. /sine/[sɨ̥ne] ‘shin’, /kata/[kḁta] ‘shoulder’).

The contrast between the closed vowels /i/ and /u/ is neutralized after /s, c, z/, and they appear as /i/[ɨ]. For example, when the verbal suffix //(r)u//(npst) attaches to an s-stem verb such as das- ‘put out’, it is realized as /das-i/[dasɨ] (put out-npst).

/i/ has three allophonic variants: the front vowels [i] and [i̞] appear after a vowel, while the central vowel [ɨ] appears in other environments (e.g. /sjooi/ [ɕoːi] ‘soy sauce’, /ki/ [kˢɨ̥] ‘tree’). /u/ does not appear in word-initial position.

/t/ and /d/ appear only before /a, e, o/.¹ /k, g/ have allophones with contain a fricative off-glide [s, z], which often appear before /i/ (e.g. /ki/[kˢɨ̥] ‘tree’). /c, z/ are affricates. /c/ is palatalized before /y/ (e.g. /cya/[tɕa] ‘Japanese tea’), while /z/ is always palatalized before /y/ and /e/ (e.g. /mesizyakusi/[mesɨdʑakɯ̥sɨ] ‘rice scoop’, /zene/[dʑene] ‘money’). The fricative /s/ is also palatalized before /y/ and /e/ (e.g. /ase/[aɕe] ‘sweat’, /bosya/[boɕa] ‘bath’). Note that the palatalization of consonants before /i/ is optional in Nita, while the consonant before /i/ is almost obligatorily palatalized in most of the other Japanese-Ryukyuan dialects (e.g. /ci/[tsɨ̥ ~ tɕɨ̥] ‘blood’). This may be due to the phonetic nature of /i/ in Nita, which is a central vowel [ɨ] after a consonant.

/h/ is realized as [ɸ] before /u/ and /i/. /n/ is realized as the dental nasal [n] in onset position, but in coda position it is a homorganic nasal, which is realized as [ɴ] in word-final position, [ŋ] before velars, [m] before bilabials, and as a nasalized vowel before vowels, fricatives, and approximants. /r/ is realized as [ɭ] before /i/ (e.g. /kiiri/[kɨːɭɨ] ‘cucumber’).

Glide phonemes may combine with other consonants to form complex onsets, as in /hwaa/ [ɸaː] ‘day’, /kwaama/ [kʷaːma] ‘car’, /kyaa/ [kʲaː] ‘fog’.

2.2 Syllable Structure and Mora

The syllable template in the Nita dialect is schematized as (O)(G)Nu₁(Nu₂)(Cd). The syllable contains an obligatory nucleus (Nu), which can be filled by vowels. The onset position (O) can be occupied by any consonant, while the coda position (Cd) cannot be occupied by the voiced obstruents /b, d, g, z/, the glides /w, y/, /m/, and the liquid /r/ in the data at hand. The glides /w, y/ can fill the G slot. /w/ precedes /k, g/ and /h/ only (e.g. /kwasi/[kʷasɨ̥ ~ kʷaɕɨ̥] ‘sweets’; /hwaa/[ɸaː] ‘daytime’). /y/ can follow any consonant but /y/ itself, and precede /a/ or /o/ (e.g. /syooi/[ɕoːi] ‘soy sauce’; /kaw-ya/[kawʲa] (buy-cond)).

The mora, as opposed to the syllable, plays a significant role in the description of the accentuation (§ 2.3). The nucleus and coda each constitute one mora, while the onset and glide are not moraic.

2.3 Word-Level Prosody

The prosodic system of Nita is a ‘pitch accent’ system (Uwano 2012). This dialect has n+1 distinctive pitch patterns for n morae words, which are distinguished by whether a fall in pitch exists or not and, if there is a fall in pitch, where it is located. Some basic data are given in Table 8.2. The acute accent mark indicates the locus of the ‘accent’; the pitch falls after an accented mora.

2.4 Intonation

For Japanese, Igarashi (2018: 189) defines an Accentual Phrase (AP) as “having a delimitative rise to high around the second mora and a subsequent gradual fall to low at the end of the phrase and as having at most one lexical pitch accent”, and further defines the Intonation Phrase (IP) as “the prosodic phrase immediately above the AP in the hierarchy within which pitch range is specified” (Igarashi 2018: 193). These definitions can also be applied to describe the intonational phrasing of Nita, and the intonational phrasing may mark a focus domain, though further research is needed.

Pitch movements at the end of the prosodic phrase contribute to the pragmatic or modal interpretation of the utterance. For instance, a rise in pitch on the final syllable of the utterance indicates a question (§ 11.2). Interaction between modal markers, discourse markers, and intonation may create various modal or pragmatic indexes, but these details have yet to be researched.

2.5 Alternations Concerning the Liquid /r/

In Nita, some (morpho)phonological processes are observed: the neutralization of closed vowels after /s, c, z/ (§ 2.1), segmental alternations in the verb morphology (§ 5), and vowel coalescence across the boundary between an adjective stem and a suffix (§ 6.1). In this section, one of the notable phonological processes, the alternations concerning the liquid /r/ as shown in (349), are described. Other phonological processes will be detailed in the following sections.

(349) a. arigo ~ aago ‘ant’

b. tori ~ too ‘bird’

c. kwaama ~ kuruma ‘car’

d. kiri ~ kyaa ‘fog’

e. eri ~ yai[jae] ‘collar’

Two pieces of evidence lead us to believe that each of the paired forms in the alternations in (349) is considered to be derived from a single underlying form, and that the form containing /r/, such as kuruma, must be the underlying form. Strong evidence of this view is found in the morphophonemic alternations in the verbal inflections as shown in (350).²

(350) a. //ar-u// → /aa/ (exist-npst)

b. //tor-u// → /too/ (take-npst)

c. //cikur-u// → /cikwaa/ (make-npst)

d. //kir-u// → /kyaa/ (cut-npst)

e. //ker-u// → /kyai/ (kick-npst)

Based on the non-past forms of the other C-verbs such as kak-u (write-npst), an allomorph of the non-past suffix must be //-u//, and the conditional forms such as /ar-ja/ (exist-cond) or the negative non-past forms such as /cikur-a-n/ (make-thm-neg.npst) show that all the stems of the verbs in (350) must end in /r/. See § 5 for more details on verb morphology.

Supporting evidence for this view is found in the relationship between the segment and the prosody. In this dialect, while the locus of fall in pitch is distinctive, the locus of the rise in pitch is predictable from the segmental conditions of the word. For example, if the first syllable of the word contains a long vowel or diphthong or it ends in /n/, the word begins with high pitch (e.g. kiiri HHH ‘cucumber’). On the other hand, if the second mora of the word contains a closed vowel (i, u) and the third mora contains a non-closed vowel (a, e, o), the rise in pitch occurs after the second mora (e.g. kobira LLH ‘calf’). The fact that kwaama appears as LLH suggests that the underlying form of this word is /kuruma/, and the second mora contains a closed vowel /u/. Note that the form kuruma is used in somewhat formal situations.³

4.1 Pronouns

Nita has personal, demonstrative, and interrogative pronouns. Non-speech-act participant reference, i.e. third person, location, and direction, is intertwined with the demonstrative system (see § 8). Here, an overview of the personal pronouns is given.

Table 8.3 shows the personal pronoun system of Nita. Pronouns of the Nita dialect are considered to inflect for number, i.e. they distinguish different word forms for different numbers (e.g. ora-∅ (1-sg) vs. ora-nci, ora-raci, and ora-yaci (1-pl)).

The first-person pronoun adan tends to be used by female speakers, while ora(-nci) is a general form, which is used by both male and female speakers.

Different forms for second-person referents are used depending on the hierarchical social relationship between the speaker and the addressee; the forms waa, waaci are used to indicate that the speaker finds themself to be of higher social status, while omai is used to indicate that the speaker considers themself to be of equal social standing with the addressee. To show the recognition that the addressee is superior on the social hierarchy, the honorific suffix -san is attached to the pronoun (e.g. omai-san (2-hon)), and it cannot be attached to waa or waaci.

The differences between the plural suffixes -nci, -ci, -raci and -jaci are not known in detail, but they have different distributions. For example, -ci attaches only to first-person adan and second-person waawa or omai. -nci attaches only to ora, and the form ora-nci may not only refer to more than two persons, but also to one person. waaci contains the plural suffix -ci, but it refers to one person only, and it cannot refer to two or more persons. Both -raci and -yaci can be attached to nouns denoting humans and animate beings. In other words, nouns denoting inanimate entities do not distinguish number. For example, in (351a), the distal demonstrative pronoun aa refers to something to eat, meaning that the plural suffix cannot be used, thus rendering the number of things to eat ambiguous. On the other hand, in (351b), the plural suffix is obligatorily attached to the demonstrative aa, which refers to someone who is not participating in the speech act, therefore, expressing plurality.

(351)

a.

{aa/*aa-raci/*aa-yaci}⸗ga	kuu-ta-i
{that/that-pl/that-pl}⸗nom	eat.thm-des-npst

‘(I) want to eat that/those.’

b.

{*aa/aa-raci/aa-yaci}⸗ga	kuu-ta
{that/that-pl/that-pl}⸗nom	eat-pst

‘Those people ate (it).’

The plural suffixes may follow the honorific suffix. For example, omai-san-ci, omai-san-raci is used to express the second person, but *omai-san-yaci is not acceptable.

4.2 Nouns

Nouns do not inflect. Cross-linguistically common inflectional categories for nouns such as number, gender, and case are either absent (in the case of gender) or present in systems other than inflection; case is marked by a clitic attaching to an NP, and plural marking is derivational, i.e. it is not obligatory for nouns.

The plural suffixes -yaci and -raci attach to nouns referring to humans and animate beings: taroo-raci (Taroo-pl), sensee-yaci (teacher-pl), hebi-raci (snake-pl).

4.3 Numerals

A numeral may be composed of a numeral root and a classifier suffix. Two usual numeral sets are as follows. In some numerals, as in (353) below, numbers greater than three or five may be expressed by numerals of Sino-Japanese origin.

(352) For general non-animate objects: huto-ci ‘1’, hutaa-ci ‘2’, mec-ci ‘3’, yoc-ci ‘4’, eci-ci ‘5’, moc-ci ‘6’, nana-ci ‘7’, jac-ci ‘8’, kokono-ci ‘9’, too ‘10’.

(353) For humans: hutoo ‘1’, hutaa-ri ‘2’, san-nin ‘3’, yottaa ‘4’, go-nin ‘5’, roku-nin ‘6’, sici-nin ‘7’, haci-nin ‘8’, ku-nin ‘9’, zii-nin ‘10’.

Numerals behave like other nominals and can head an NP, but they can also float (e.g. ringo hutoci kuu-ta (apple one eat-pst) ‘I ate one apple’).

4.4 Formal Nouns

Nita has some formal nouns, which have undergone some grammaticalization. Formal nouns cannot be used alone, and the modifier always precedes them. Some of them have entirely lost their lexical meaning and are now pure grammatical markers, while others still retain lexical meaning. For example, eci ~ yaci, which may have originated from yaci ‘guy’, can be used not only for humans and animate beings but also for inanimate things or events as in (354). On the other hand, sii, which may be derived from sii (< *syuu ‘people’, from Chinese zhòng), can be used only for persons (355).

(354)

kaee⸗ga	negee	yaci	mi-ta
frog⸗nom	run.away.npst	YACI	see-pst

‘(I) saw a frog run away.’

(355)

moo	ano	sii	ene+kake-cyot-te⸗da⸗yo
already	that	SII.top	leave.inf+begin-cont-seq⸗cop.npst⸗dsc

‘That person is about to leave.’⁴

4.5 Compound Nouns

In Nita, like other Japanese-Ryukyuan dialects, compounding is very common in nominal (and verbal) morphology, and rendaku (sequential voicing) is also common. For example, the second element of each compound noun of (356a–c) is subject to rendaku: hara ‘stomach’ → bara, syakusi ‘ladle’ → zyakusi, kwaa ‘chestnut’ → gwaa. The plus sign “+” here indicates the boundary between elements of a compound.

(356)

a.

waki+bara

side+stomac

‘flank’

b.

mesi+zyakusi

rice+ladle

‘rice scoop’

c.

ega+gwaa

burr+chestnut

‘chestnut in its burr’

5.1 Inflectional Morphology

Inflection covers the categories of tense, mood and polarity,⁶ but also encodes differences in the morpho-syntactic status of forms, i.e. finite or non-finite (see Table 8.5).

The suffix-initial /r/ in -ru (npst), -re (imp), -runa (proh) and -rya(a) (cond) is omitted when these suffixes are attached to the C-verb stem (i.e. //kak-re// → /kak-e/ write-imp).

-oo, one of the allomorphs of the inferential suffix, follows the V-verbs and come and do verbs. In this case, the stem-final vowels /i/ and /e/, if present, alternate with /y/ (i.e. //ake-oo// → /aky-oo/ ‘will open’; cf. /k-oo/ ‘will come’). On the other hand, the allomorph -aa follows C-verbs (i.e. kak-aa ‘(Someone) will write (something)’).

The infinitive forms participate in compounding as in (357a,b).

(357)

a.

nahayasi

na+hayas-i-∅

greens+cut-thm-inf

‘knife for cutting greens’

b.

enesagare!

en-i-∅+sagar-re

leave-thm-inf+step.back-imp

‘Go away!’

Note that there are negative forms such as kaka-sen ‘do not write’, which are derived from kak-i-∅⸗wa se-n (write-thm-inf⸗top/cntr do-neg.npst).

5.2 Derivational Morphology

Non-class changing verbal derivation includes the causative (-sase-), passive (-ra(r)e-), potential (-e-, -ra(r)e-), aspectual (-cyor-), honorific (-ra(r)e-, -nahar-, -syar-, -cyar-). The different honorification suffixes may be used depending on the hierarchical social relationship between the speaker and the subject. The speaker uses the suffix -cyar- to show a moderate degree of respect to the subject, while, to show a higher degree of respect, the speaker must use the other suffixes.⁷

The polite marker -mas- can often be observed in formal speech, and its inflection may be restricted; only the non-past indicative form -mas-i as in kak-i-mas-i (write-thm-pol-npst) ‘(I) will write it’ and the past habitual form -mas-yot-ta as in ii-mas-yot-ta (say.thm-pol-hab-pst) ‘(I) used to say it’ have been observed.

5.3 Existential and Copular Verbs

In Nita, there are two existential verbs: or- and ar-. The former is used for animate subjects, while the latter is for inanimate subjects. Their inflectional patterns are almost identical to that of r-verbs, except that the negative adjective na- is used to express the non-existence of an inanimate subject instead of *ar-an. Table 8.6 shows the paradigm of the copular verb ⸗dar-. In addition to the forms in Table 8.6, the copula has a special polite form ⸗des-.

The n-series forms are used only for nominal adjectives (see § 6.2), while the d-series forms are used for all the types of nominals.

In negation, the analytical forms ⸗da na- ~ ⸗zya na- (//de⸗wa na// (cop.seq⸗top neg)), ⸗nya na- (//ni⸗wa na// (cop.adv⸗top neg)) and ⸗n na- (//ni na// (cop.adv⸗top.neg)) are used.

⸗da, the non-past indicative form of the d-series, does not precede a noun (i.e. *otoko⸗da huto (male⸗cop.npst person) ‘male person’). Instead, the genitive particle ⸗no/ga is used when an NP modifies another NP (e.g. otoko⸗no huto (male⸗gen person) ‘male person’).

6.1 Verbal Adjectives

Tables 8.7 and 8.8 give the paradigm of the verbal adjective haya- ‘fast/early’.⁸

Vowel coalescence optionally takes place across a boundary between an adjective stem and the indicative non-past suffix -i. As shown in (358), /ai/ and /oi/ alternate with /e(e)/, and /ui/ alternates with /i(i)/ via vowel coalescence, which is optional.

(358) a. //taka-i// (high-npst) → takai ~ take(e)

b. //too-i// (far-npst) → tooi ~ toe(e)

c. //hiku-i// (low-npst) → hikui ~ hiki(i)

Infinitive forms participate in adverbial expressions such as haya-∅ koto and haya-∅⸗n(i), and such adverbial expressions are used to form so-called light-verb constructions such as haya-∅ koto saa ‘hurry up (lit. do something in haste)’ or haya-∅⸗n(i) naa ‘become fast(er)’.

The complex verbal phrase is used to negate the verbal adjective. The adjective inflects as an infinitive form and the auxiliary verb is the negative verbal adjective na- (e.g. haya-∅ na-i (fast/early-inf neg-npst) ‘not fast/early’).

6.2 Nominal Adjectives

Nominal adjectives do not inflect and may take a copula to indicate tense, mood, polarity and finiteness. Moreover, as shown in § 5.3, nominal adjectives may take special n-series copular forms as in (359). Note that some nominal adjective roots may occur in a special construction X⸗ga aa, where X may be a noun or a nominal adjective as shown in (360).

(359)

{genki⸗na/genki⸗da}⸗ka

fine⸗cop.npst⸗q

‘(Are you) fine?’

(360)

genki⸗ga	aa
fine⸗nom	exist.npst

‘(S/he) is fine.’

7.1 Nominalization and Nominalizers

In Nita, three nominalizers, ⸗∅, ⸗ga and ⸗no, have been identified. In (361), the phrase yai-ta⸗∅ serves as an object. The slot filled by ∅ can be filled by the nominalizer ⸗no or formal nouns such as bun or yaci. Younger speakers tend to fill the slot with such overt forms.

(361)

{yai-ta⸗∅	/	yai-ta⸗no	/	yai-ta yaci	/	yai-ta bun}⸗o	kuu-ta
{bake-pst⸗nmlz	/	bake-pst⸗nmlz	/	bake-pst YACI	/	bake-pst BUN}⸗acc	eat-pst

‘I ate the baked one.’

The nominalizer ⸗∅ is often used in an event nominalization as observed in (362a). ⸗∅ cannot be used for a referential nominalization which refers to a person as in (362b). In contrast, it can be used in referential nominalizations which refers to a non-human referent as in (363a,b).

(362)

a.

kyonen	aa-ta⸗∅⸗o	oboe-cyoo
last.year	meet-pst⸗nmlz⸗acc	remember-cont.npst

‘Do you remember that we met last year?’

b.

kyonen	{*aa-ta⸗∅	/	aa-ta sii}	oboe-cyoo
last.year	{meet-pst⸗nmlz	/	meet-pst SII}	remember-cont.npst

‘Do you remember the person you met last year?’

(363)

a.

sara⸗ne	at-ta⸗∅⸗ga	naanat-ta
plate⸗dat	exist-pst⸗nmlz⸗nom	disappear-pst

‘What was on the plate was gone.’

b.

sara⸗ne	at-ta⸗∅⸗o	ora⸗ga	kuu-ta
plate⸗dat	exist-pst⸗nmlz⸗acc	1st⸗nom	eat-pst

‘I ate what was on the plate.’

7.2 Other Class-Changing Derivations

Verbal adjective stems are derived from verb roots with the desiderative suffix -ta- (e.g. kuu-ta-karaa⸗zi (eat.thm-des-infr⸗mod) ‘(You) will want to eat’). Several verb stems which are derived from nouns have been identified (e.g. kyaa-gom- (fog-vlz) ‘become foggy’).

9.1 Basic Structure

A nominal phrase (NP) consists of the head and (optionally) a modifier, which may be filled by a noun plus a genitive particle ⸗no/⸗ga, an adjective, an adnominal, or an adnominal clause. An argument NP is followed by case and other role markers.

9.2 Case Marking

In Nita, the case particles listed in Table 8.10 have been identified.

As shown in Table 8.10, the same case forms are used as nominative and genitive markers. For convenience, I use the term ‘nominative’ (nom) when the marker is attached to an NP that serves as an argument of the predicate, and ‘genitive’ (gen) when the marker is attached to an NP that serves as a modifier of a larger NP. As in (366a), when the relative social rank of the NP to which the marker is attached is high, ⸗no is used.¹⁰ In the case of the nominative, only ⸗ga is used in the main clause, while ⸗no can be used in adnominal clauses. On the other hand, when the relative social rank of the NP is not higher than the speaker’s, the choice between the two markers is based on the animacy-definiteness of the NP. In the case of the genitive, when the NP to which the marker is attached refers to human beings, both ⸗ga and ⸗no can be used (366b), while only ⸗no can be used when the NP refers to animals or inanimate things (366c).

(366)

a.

sense⸗{no/*ga}	tenugui
teacher⸗gen	towel

‘a towel of our teacher’

b.

taroo⸗{no/ga}	tenugui
Taro⸗gen	towel

‘Taro’s towel’

c.

cikue⸗{no/*ga}	nezi
desk⸗gen	screw

‘desk screw’

As suggested by (367a–c), the relative rank in the animacy hierarchy of the subject and object may affect the choice to use or not use the overt accusative marker ⸗o, though further investigation is needed.¹¹

(367)

a.

ziroo⸗ga	{*omae	/	omae⸗o}	mi-cyoo
ziroo⸗nom	{2	/	2⸗acc}	look.at-cont.npst

‘Ziroo is looking at you.’

b.

ziroo⸗ga	{inu	/	inu⸗o}	mi-cyoo
ziroo⸗nom	{dog	/	dog⸗acc}	look.at-cont.npst

‘Ziroo is looking at a dog.’

c.

ora⸗wa	{omae	/	omae⸗o}	mi-cyoo
1⸗top	{2	/	2⸗acc}	look.at-cont.npst

‘I am looking at you.’

9.3 Other Role Markers

So far, the following role markers have been identified: ⸗mo (add ‘also, even’), ⸗dari (add ‘even’), ⸗kurai (exm), ⸗demo (exm), ⸗nato (exm), ⸗doma(a) (exm), ⸗sika (lmtd ‘only’), ⸗dake (lmtd ‘only’), ⸗hodo (lmtd ‘only, at least’), ⸗bakka(a) (lmtd ‘just’), ⸗wate (dist ‘at a time, each’), and ⸗wa (cntr). See the sample text for concrete examples of ⸗kurai, ⸗sika (398), and ⸗dake (400).

In the data at hand, there are some cases where a role marker follows the case particle as in (368), while there are no examples where it precedes the case particle.

(368)

tanzyoobi⸗ni⸗doma	kaet-tara
birthday⸗dat⸗exm	return-cond

‘Come home at least on your birthday.’

The same form as the contrastive ⸗wa is also used as a topic marker (see § 11.7). Note that the topic (and contrastive) marker ⸗wa and the additive markers always replace the nominative and accusative markers (e.g. ora⸗wa not *ora⸗ga⸗wa). Further research will be needed to clarify the co-occurrence restrictions of case and other role markers.

10.1 Verbal Predicate

A verbal predicate consists of a lexical verbal root (either a verb or a verbal adjective) and optionally an auxiliary verb, which dilutes or changes the lexical meaning and expresses a grammatical meaning. In a complex verbal predicate, a lexical verb inflects as the sequential converb -te, and the auxiliary verb inflects for tense, mood, and polarity. The auxiliary verb constructions in (369) have been identified. Benefactive auxiliary verbs derive from verbs of giving (yar-, age- ‘give others’, gos- ‘give us’) or receiving (moraw-, maw-). Aspectual auxiliary verbs will be described in § 11.5.1. In addition to (369), -te mi- is used to designate the speaker’s intention to try (deontic modality).

(369) a. Benefactive: -te yar-, -te age-, -te gos- (giving of a favor), -te moraw-, -te maw- (receiving of a favor)

b. Aspectual: -te ar- ~ -cyar- (resultative), -te ok- ~ -cyok- (preparatory), -te maw- ~ -cyaw- (completive)

10.2 Nominal Predicate

A nominal predicate consists of an NP as the head of the predicate and a copular verb, which inflects for tense and mood as would a general verb (see § 5.3 for the morphology of copular verbs).

A nominalized phrase (or clause) with the nominalizer ⸗∅ can be followed by a copular verb as in (370), which is presented as new information for the addressee (see also § 11.2 and § 11.7).

(370)

taroo⸗wa	sara⸗o	wat-ta⸗∅⸗da⸗yo
taroo⸗top	plate⸗acc	break-pst⸗nmlz⸗cop.npst⸗dsc

‘Taro broke a plate.’

11.1 Alignment System and Non-canonical Case-Marking

The alignment system of Nita is a nominative-accusative system, in which S/A are obligatorily marked with a nominative case marker ⸗ga/⸗no and P is optionally marked with accusative ⸗o.¹² In a clause whose predicate is a ditransitive verb (e.g. yar- ‘give others’, okur- ‘send’), the recipient or goal is marked with dative case as in (371), and the causee agent is also marked with the dative case in a ditransitive clause derived by the causativization (see § 11.4).

(371)

taroo⸗wa	otooto⸗ne	waawa⸗no	ie⸗o	yat-ta
Taroo⸗top	brother⸗dat	refl⸗gen	house⸗acc	give-pst

‘Taro gave his house to his younger brother.’

Nita has a double nominative construction in which the predicate takes two nominative arguments (372). In this construction, the first NP is marked with ⸗wa (top), and the second NP with ⸗ga (nom).

(372)

kome⸗wa	nita⸗ga	ma-i
rice⸗top	Nita⸗nom	good.taste-npst

‘(For) rice, Nita is (a) good (place).’

Nita also has a transitive adjectival construction in which the predicate takes an experiencer and stimulus (373). In this construction, the two NPs are canonically nominative-marked as in the double nominative construction, while the second NP can be marked with ⸗ni (dat). Such constructions can be formed with predicates such as suki⸗dar- (love⸗cop) or suk- (like).

(373)

adan⸗wa	ni-i-∅+mono⸗{ga/ni}	sik-a-n
1⸗top	sew-thm-inf+MONO⸗{nom/dat}	like-thm-neg.npst

‘I do not like sewing.’

11.2 Sentence Type

Based on morpho-syntactic and prosodic properties, three distinct sentence types are identified: declaratives, interrogatives, and imperatives. This section provides an overview of interrogatives and imperatives.

Interrogative sentences divide into two subtypes: polar and content interrogatives. In Nita, both polar and content questions are marked with question particles such as ⸗ka. Furthermore, rising intonation in sentence-final position indicates that a question is being asked (374, 375). The final rising intonation is also observed when the question marker ⸗ka is used. Further research is needed to clarify the relationship between question markers and intonation.

(374)

zisin⸗ga	ar-ta(⸗no)	{sit-cyor-u⸗ka⸗ne(↘)	/	sit-cyor-u(↗)}
earthquake⸗nom	exist-pst(⸗nmlz)	{know-asp-npst⸗q⸗dsc	/	know-asp-npst}

‘Did you know that there was an earthquake?’

(375)

korya	daa⸗no	{tenogui⸗ka⸗ne(↘)	/	tenugui⸗da(↗)}
this.top	who⸗gen	{towel⸗q⸗dsc	/	towel⸗cop}

‘Whose towel is this?’

Imperatives are commands addressed to a second person. Nita has the following three forms dedicated to expressing commands: (a) the imperative inflection (oki-re or oki-tae(⸗yo)), (b) the imperative forms of the honorific suffixes, as in kak-i-naha-i (write-thm-hon-imp) ‘Please write it’, and (c) the imperative forms of the benefactive auxiliary verbs, as in noo-te gos-e (sew-seq ben-imp) ‘Please sew it’. In addition, there are some conventionalized ways of expressing commands pragmatically, and these are declarative in morphological terms: (d) the sequential form, as in kak-a-sikoni oi-te (write-thm-neg.seq put-seq) ‘Don’t write it down, just let it be’, (e) the non-past form plus copula construction, as in kak-u⸗∅⸗da (write-npst⸗nmlz⸗cop) ‘Write it!’, (f) the past form, as in sore goi-ta (it give-pst) ‘Please give it (to me)’.

Prohibitives (negative commands) are marked with the inflectional suffix -runa.

11.3 Possession

In the possessive construction, the possessor is marked with dative case ⸗ni and the possessed with nominative case ⸗ga. In this construction, the existential verbs ar- and or- are used (376).

(376)

oci⸗nyaa	eno-ga	oo⸗yo
(My) family⸗dat.top	dog⸗nom	exist.npst⸗dsc

‘We have a dog.’

When the possessed is inanimate and alienable, mot- ‘have, hold’ may be used, and in this case, the possessor is marked with the nominative case and the possessed with the accusative case.

11.4 Valency Changing

11.5 Aspect and Tense

11.6 Mood and Modality

In Nita, two opposing types of mood have been identified, and they are marked with mood inflectional suffixes: declarative and imperative. See § 5.1 and § 11.2 for each inflection and the relationship with sentence type.

Intentional modality is expressed by the non-past form plus the modal marker ⸗ka, or the discourse marker ⸗wa or ⸗jo. The intentional form is not used in isolation to express volition. It is always followed by ⸗to (o)mow- (⸗quot think) as in the second sentence of the sample text.

The intentional form of a verb may be followed by the modal marker ⸗koi, which comes from the imperative form of the ‘come’-verb, or ⸗ya, and it expresses hortativity as in (386).

(386)

asita⸗mo	koko⸗e	k-oo⸗koi
tomorrow⸗add	here⸗all	come-int⸗hor

‘Let’s come back here together tomorrow.’

Inferential (epistemic) modality is expressed by intentional forms plus the modal markers ⸗zo/⸗zi (387a). Note that the formal noun sikoo(⸗da) is used to express inferential modality as in (387b) and that the inferential form of the copula ⸗dar-aa is also used for the inferential as in (387a,b). Additionally, there are nominal adjective constructions such as ge(⸗na), ya(a)(⸗na), saa(⸗na), which express evidential modalities, although the difference between these constructions is still unclear.

(387)

a.

sorosoro	{k-oo⸗zo⸗naa	/	kwaa⸗dar-aa}
soon	{come-int⸗mod⸗dsc	/	come.npst⸗cop-infr}

‘(He) will be here soon.’

b.

asita⸗wa	{huu sikoo⸗da	/	huu⸗dar-aa}
tomorrow⸗top	{rain.npst SIKOO⸗cop.npst	/	rain.npst⸗cop-infr}

‘It will rain tomorrow.’

Some kinds of complex predications which express deontic modalities have been found (388), but the details are not yet clear.

(388)

haya	benkyo+suu⸗ga	ee⸗ga⸗na
soon	study+do.npst⸗nom	good.npst⸗mod⸗dsc

‘The sooner you study, the better.’

11.7 Information Structure and Its Formal Encodings: Topic and Focus

In principle, a topicalized argument is marked with ⸗wa, and it appears in sentence-initial position (e.g. the subject taroo is topicalized in (378a), while the object kono ie is topicalized in (379a)). However, as in (355), the topicalized argument may appear without any particle.

In Nita, there is no dedicated morphological focus marker. A focus is marked with a cleft construction ‘X⸗wa Y⸗da’, in which Y is the focus (389).

(389)

akamboo⸗ga	nai-cyoo⸗wa	o-naka⸗ga	hec-cyoo⸗ken⸗da⸗wa.
baby⸗nom	cry-cont⸗top	pol-stomach⸗nom	decrease-cont⸗csl⸗cop.npst⸗mod

‘(I think that) It is because s/he is hungry that the baby is crying.’

The intonation must also have something to do with information structure, but this has not been investigated in detail.

12.1 Coordination

In Nita, syntactic clausal coordination is not prominent, and clause-chaining constructions are used instead as a coordination strategy (§ 12.3). Only constructions with the conjunctive markers ⸗si, ⸗damo and ⸗ne (390a,b) are identified as coordination constructions.

(390)

a.

eno⸗mo	oo⸗si	neko⸗mo	oo(⸗si)
dog⸗add	exist.npst⸗and	cat⸗add	exist.npst(⸗and)

‘There are dogs, and (there are) cats.’

b.

eno⸗wa	oo⸗{damo/ne}	neko⸗wa	or-an
dog⸗cont	exist.npst⸗but	cat⸗cont	exist-neg.npst

‘There are dogs, but there are cats.’

12.2 Subordination

There are three types of subordinate clause: complement clauses, adnominal clauses, and adverbial clauses.

Complement clauses divide into two subtypes: noun clauses and quotative clauses. Noun clauses are formed by attaching a nominalizer, and they serve as arguments (see § 7). Quotative clauses are formed by being directly followed by verbs of saying or thinking, or by attaching the quotative markers ⸗to and ⸗tte. See the sample text for concrete examples of quotative clauses with ⸗to (394) and ⸗tte (402) and without any marker (394 and 402).

An adnominal clause precedes a head noun without a relativizer such as a relative pronoun (391). Verbals in adnominal clauses inflect as the unmarked forms (-ru, -n, -ta, -i, etc.).

(391)

[oci⸗ne	at-ta]	sara	zyanzyane	mee-da
[my house⸗dat	exist-pst]	plates	many	break-pst

‘(I) broke many plates in my house.’

Adverbial clauses are most usually headed by converbs (392).

(392)

cyonbo	ake-ryaa	sizisi⸗n	naa⸗yo
a little	open-cond	cool⸗cop.adv	become.npst⸗dsc

‘If you open (the windows) a little, it’ll be cooler.’

The other adverbial subordination strategy available is the use of a finite form followed by the formal noun nakai ‘interval’, and the nominalizer ⸗∅ followed by a conditional form of copular verb such as saa⸗∅⸗nara (do.npst⸗nmlz⸗cop.cond).

12.3 Clause-Chaining

In the clause-chaining construction, a series of non-finite (converbal) clauses follow each other and only the final verb is finite carrying the TAM markers. Clause-chaining constructions are close to adverbial subordinate structures in that they connect non-finite clauses, and they are also close to coordinated structures in that they are used as a coordination strategy. Chain-medial clauses are headed by a sequential converb. See the first sentence in the appendix for a concrete example.

12.4 Insubordination

Nita exhibits an insubordination process whereby a non-finite clause is used as a main clause. For example, in the third sentence in the appendix, the causal marker ⸗ken behaves as a discourse marker, and it may express ‘I envy her because I never return to my parents’ home for hange domari.’