2.1.1 Vowels

Shiiba has five vowels (/i, e, a, o, u/). The vowel /u/ is a less rounded [u̜], but it is transcribed as [u] throughout this chapter. The vowel /e/ is realized as [je] syllable-initially (/mo.to.e/ [motoje] ‘main family’, /e.no.ha/ [jenoha] ‘trout’). The vowel /o/ tends to be realized as [wo] after a vowel, especially after a high vowel (/kao/ [ka.wo]~[ka.o] ‘face’, /io/ [iwo] ‘fish’, /suo/ [su.wo] ‘nest:acc’, etc.). A long vowel is phonemically interpreted as a sequence of two identical vowels. Diphthongs are scarce in roots (e.g. /kaiko/ ‘silkworm’, /kauya/ ‘river’, /oiko/ ‘nephew’, etc.). Diphthongs arising from morpheme combinations regularly change to long vowels by a vowel fusion rule (see § 2.4).

View Full Size

Figure 10.1

Shiiba village

• Download Figure
• Download figure as PowerPoint slide

2.1.2 Consonants and Glides

Shiiba has twelve consonants (/(p), t, k, b, d, g, s, h, z, m, n, r/) and two glides (/w, y/). The labial stop /p/ never occurs in native roots except in geminates (e.g. /happa/ ‘leaf’). /r/ is a tap [ɾ], and does not occur initially in native roots. The nasal /n/ is a homorganic nasal whose place and manner features are determined by the segment that follows it (e.g. /hon/ [hoɴ] ‘book’, /zunbyaa/ [d͡zumb^jaː] ‘plenty’, /binta/ [binta] ‘head’, /senso/ [sez̃so] ‘ancestor’, /bebenko/ [bebeŋko] ‘calf’).

Systematic allophonic variations are observed in alveolar obstruents (Table 10.1).

2.4.1 Sequential Voicing

Sequential voicing applies to compounding, changing the initial voiceless onset C of the medial base to its voiced counterpart, as in tyaag̲wasi ‘tea snack’ (//tya// ‘tea’ + //k̲wasi// ‘snack’), where the underlined segment undergoes this process. Examples such as kusob̲yaa ‘fly of some kind’ (//kuso// ‘feces’ + //h̲wyaa// ‘fly’) suggest that /h/ and /b/ form a natural class, which is due to the fact that /h/ corresponds to proto-Japonic *p (see § 2.2).

2.4.2 T-suffixation

The past-tense suffix -ta, the sequential converbal suffix -te and the completive suffix -tor- share exactly the same morphophonological feature. They will be called ‘T-suffixes’ henceforth.³ Table 10.2 illustrates the morphophonological pattern of T-suffixation with a T-suffix -ta (past).

A voicing rule converts the /t/ of T-suffixes to its voiced counterpart. A C-to-V alternation rule turns the stem-final C /s, k, g/ to /i/ and /b, m, w/ to /u/.

2.4.3 Vowel Fusion Rule

A vowel fusion rule (VFR) applies in three major morphological contexts, T-suffixation (§ 2.4.2), dative case cliticization and verbal-adjectival inflection.⁴ In dative case cliticization, the dative case particle ⸗ni may yield a diphthong in the middle of its derivation, with deletion of its initial /n/ depending on the stem-final segment (e.g. //koko⸗ni// (this.place⸗dat) → kokoi). Verbal-adjectival inflection (§ 6.1) also yields a diphthong in non-past tense, as in //aka-i// (red-npst).

The derived diphthongs then undergo VFR. (459) illustrates VFR which derives (w)yaa from /ai/. The bracketed /w/ is deleted if the resulting form would lead to an impermissible CG cluster. Whereas (459a) surfaces as /wyaata/ (CGVV.CV) with no deletion of /w/, (459b) surfaces as /kyaata/ (CGVVCV) instead of */kwyaata/ (CGGVVCV). In (459c), /hwyaata/ does not undergo deletion despite the impermissible CCG cluster (see § 2.2 and footnote 1 for a historical account).

(459) a. //ak-ta// (open-pst) → aita → VFR (ai → wyaa) → wyaata ‘opened’

b. //kak-ta// (write-pst) → kaita → VFR (ai → wyaa) → *kwyaata → kyaata ‘wrote’

c. //hak-ta// (vomit-pst) → haita → VFR (ai → wyaa) → hwyaata ‘vomited’

Another VFR derives /Gee/ from //oi// (460), where G is an unspecified glide slot which is filled by /w/ if possible, and otherwise by /y/ to derive a permissible phonotactic structure. If neither is permissible, then G is deleted altogether by a general strategy of cluster reduction.

(460) a. //too-i// (distant-npst) → tooi → VFR (//oi// → Gee) → towee ‘be distant’

b. //siro-i// (white-npst) → siroi → VFR → siryee ‘white’

c. //igok-ta// (move-pst) → igoita → VFR → igeeta (*igweeta, *igyeeta) ‘moved’

Other VFRs derive /yee/ from //ei// (e.g. //kes-ta// (erase-pst) → keita → kyeeta ‘erased’), /yuu/ from //eu// (e.g. //hute-u// (throw.away-int)→ huteu → hutyuu ‘will throw away’; see Table 10.7), /yuu/ from //iu// (//iw-ta// (say-pst) → iuta → yuuta ‘said’), /(w)ii/ from //ui// (e.g. /zuru-i// (cunning-npst) → *zurwii → zurii ‘be cunning’), /oo/ from //au// (e.g. //kaw-ta// (buy-pst) → kauta → koota ‘bought’) and /uu/ from //ou// (e.g. //ow-ta// (chase-pst) → outa → uuta ‘chased’).

5.2.1 Finite Inflection

The finite-inflectional paradigms of C-final and V-final stems are illustrated in Tables 10.6 and 10.7 respectively. The underlying structure of each form is indicated to the right of the italicized surface form, and the inflectional affix is indicated by [ ].

An R-final stem such as miR- ‘look at’ inflects as a C-final stem in some environments, specifically in the imperative (e.g. mir-e //miR-re//) and in cases where the C-final stem takes the /a/ thematic form, as in mir-oo //miR-a-u// (intentional). It inflects as a V-final stem elsewhere, as in mi-ta //miR-ta// (past).

5.2.2 Non-finite Inflection

Non-finite inflection is found with converbal forms, which inflect to express various adverbial or adsentential clause relations.

The sequential converb heads a chained clause (of the clause chaining structure; § 12) or the dependent component of a complex predicate (§ 9.1.1). The conditional converb heads an adverbial subordinate clause which designates the ‘if/when’ relation, or may end a sentence to function with imperative force (§ 11.1). The simultaneous converb heads an adverbial subordinate clause which designates the ‘while/during’ relation. The purposive converb heads an adverbial subordinate clause which designates the ‘in order to’ relation, always co-occurring with a directional verb such as iku ‘go’ or kuru ‘come’, etc.

The negative sequential forms -zi and -nde both head adsentential clauses (461a). They differ with respect to whether they can serve as the lexical verb component of a complex predicate (461b).⁸

(461)

a.

nan⸗mo	{kw-a-zi/kw-a-nde}	katar-i-ot-ta⸗ga
what⸗add	{eat-thm-neg.seq/eat-thm-neg.seq}	talk-thm-hbt-pst⸗sfp

‘They would talk without eating anything’ (lit. Eating nothing, they would talk.)

b.

nan⸗mo	{kw-a-zi/*kw-a-nde}	or-u⸗ga
what⸗add	{eat-thm-neg.seq/eat-thm-neg.seq}	prog-npst⸗sfp

‘They are eating nothing.’

9.1.1 Coverb Construction

A verbal predicate may contain a single verb root or may consist of two verb roots, forming a coverb construction (Table 10.13). From the perspective of wordhood, a coverb construction may be a one-word construction or phrasal. From the perspective of function, a coverb construction may be symmetrical (with two lexical roots) or asymmetrical (with a lexical root and a grammatical root).

(469)

a.

akago⸗no	[nak-i-orab-u]	koe.
baby⸗nom2	cry-thm-scream-npst	voice

‘the sound of a baby [crying out]’ (Lexical compound)

b.

akago⸗no	[nak-i-tuduku-ru]	koe.
baby⸗nom2	cry-thm-keep-npst	voice

‘(I) the sound of a baby [keeping crying]’ (Syntactic compound)

c.

io⸗oba	[yaa-te	kuta]⸗wai.
fish⸗acc	[grill-seq	eat-pst]⸗sfp

‘(I) [grilled and ate] fish.’ (Serial Verb Construction)

d.

io⸗oba	[yaa-te	kure-n]⸗no?
fish⸗acc	[grill-seq	ben-neg.npst]⸗q

‘Why not [grill fish for me]?’ (Auxiliary Verb Construction)

In both lexical and syntactic compounds, the first stem is /i/ thematic if it is C-final (e.g. /naki/ in nakiorabu in (469a)). A lexical compound differs from a syntactic compound in the functional asymmetry of the two stems in the compound structure: in a lexical compound, both stems are lexical, while in the syntactic compounds the second stem functions as a grammatical marker and is thus called a grammatical as opposed to lexical stem (e.g. /tudukE/ as a durative aspectual marker in (469b)). The two stems in a lexical compound form a single lexical item and thus a single stem nucleus, to which derivational affixes may attach (e.g. nak-i-orab-ase-ta (cry-scream-caus-pst) ‘made (someone) cry out’).¹³ Syntactic compound structure and root-affix structure may often be ambiguous due to grammaticalization processes where the syntactic compound structure is a typical source structure for an affix to develop. See § 11.6.2 for grammaticalization of aspect marking and § 11.5 for grammaticalization of potential marking.

Like the distinction between lexical and syntactic compounds, the difference between the Serial Verb Construction (SVC) and the Auxiliary Verb Construction (AVC) is that the two words in a SVC are lexical verbs while the second word in an AVC is an auxiliary. A SVC is a lexicalized pair of verbs like yaate kuu ‘grill and eat’, sariite kaeru ‘return by foot’, etc., in which the two verbs share their arguments. The AVC is used for aspectual (see § 11.6.2), directional (e.g. (472)), modal (e.g. yaate simoota ‘grilled against my intention’), benefactive (e.g. (469d)), and many more constructions which are not covered in this chapter.

9.1.2 Infinitival Phrase Construction

The /i/ thematic stem form of a C-final verb (§ 5.1) and the bare form (athematic form) of a V-final verb may function like a word, occurring as the complement of a number of complex predicate structures, even though they are uninflected. This exceptional, uninflected verb form is called the infinitive form, and the complex predicates in which the infinitive form occurs as a complement are called infinitival phrase constructions, illustrated in (470).

(470)

a.

oyano	yuu	kotaa	[kikyaa	sen].
oya⸗no	iw-ru	koto⸗wa	[kik-i⸗wa	se-n]
parent⸗gen2	say-npst	thing⸗top	[hear-thm⸗top	lv-neg.npst]

‘(Children) [never listen to] what their parents say.’ (Light Verb Construction)

b.

utino	konyaa	sono	honwa	[yomyaa	en].
uti⸗no	ko⸗ni⸗wa	sono	hon⸗wa	[yom-i⸗wa	yE-n]
our.home⸗gen2	child⸗dat⸗top	that	book⸗top	[read-thm⸗top	pot-neg.npst]

‘Our child [cannot read] that book.’ (Potential; see§ 11.5)

11.2.1 Alignment

Shiiba has a nominative-accusative alignment system where S/A is marked by the nominative case marker ⸗ga/⸗no and P is marked by the accusative case marker ⸗ba, ⸗oba or ⸗o. The differential marking of A/S and P is discussed in the following sections. In a ditransitive sentence, E (goal/recipient as an extended core argument) is marked with the dative case ⸗ni, and this holds true for a ditranstive sentence derived by causativization (see § 11.4.1 for causativization).

(471)

a.

tontyan⸗ga	modot-te	ki-ta.
father⸗nom1	return-seq	endo-pst

‘My father came back.’ (Intransitive)

b.

tontyan⸗ga	kantyoo⸗oba	utikoryeeta.
father⸗nom1	deer⸗acc	shoot.to.death-pst

‘My father shot the deer to death.’ (Transitive)

c.

tontyan⸗ga	ore⸗ni	hon⸗ba	kure-ta.
father⸗nom1	1sg⸗dat	book⸗acc	give-pst

‘My father gave me a book.’ (Ditransitive)

11.2.2 Differential Subject Marking

The choice of the two nominative markers is determined by two factors: animacy and transitivity. With respect to animacy, A/S at the higher end of the Animacy Hierarchy opts for ⸗ga (472a), while A/S at the lower end opts for ⸗no (472b). A/S at the middle of the hierarchy, especially human and domesticated animal nouns, may be marked either way (472c).

(472)

a.

tontyan⸗ga	modot-te	ki-ta.
father⸗nom1	return-seq	endo-pst

‘My father came back.’ (Animate human)

b.

basu⸗no	modot-te	ki-ta.
bus⸗nom2	return-seq	endo-pst

‘The bus came back.’ (Inanimate)

c.

inu{⸗ga/⸗no}	modot-te	ki-ta.
dog{⸗nom1/⸗nom2}	return-seq	endo-pst

‘The dog came back.’ (Animate non-human)

With respect to transitivity, A is more prone to take ⸗ga than S when the animacy of the NP in question is identical. Compare (472c) and (473), both of which have an animate non-human noun.

(473)

inu⸗ga	tyoo⸗o	uu-te	tob-i-agat-ta.
dog⸗nom1	butterfly⸗acc	chase-seq	jump-thm-rise-pst

‘The dog chased the butterfly and jumped.’ (A: animate non-human)

11.2.3 Differential Object Marking

While non-topic S/A are almost obligatorily case-marked, non-topic P may often be left unmarked. It is still unclear what determines the choice of the three accusative markers (⸗ba, ⸗oba and ⸗o). With regard to the choice between overt marking vs. zero marking of P, two factors play a crucial role: animacy and adjacency of P and V.

With respect to animacy, it is not the absolute animacy of P (i.e. the animacy of P itself), but the relative animacy of A and P, which affects overt vs. zero marking of P. As illustrated in (474a), zero marking is most common when the animacy of A is higher than that of P (A > P), which is a prototypical transitive situation with A having more control over P. The double parentheses in (474a) indicate that zero marking is more common than overt marking. By contrast, overt marking is obligatory if the animacy of P outranks that of A (A < P) (474b), a less likely transitive situation. Overt marking of P is varied among speakers when the animacy of A and P is equal (A = P) (474c).

(474)

a.

kodomo⸗ga	kantyoo((⸗oba))	mi-tor-u.
child⸗nom1	deer((⸗acc))	look.at-cpl-npst

‘The child is looking at a deer.’ (A: animate human > P: animate non-human)

b.

kantyoo⸗ga	kodomo⸗oba	mi-tor-u.
deer⸗nom1	child⸗acc	look.at-cpl-npst

‘The deer is looking at a child.’ (A: animate non-human < P: animate human)

c.

sensee⸗ga	kodomo(⸗oba)	mi-tor-u.
teacher⸗nom1	child(⸗acc)	look.at-cpl-npst

‘The teacher is looking at a child.’ (A: animate human = P: animate human)

The adjacency of P to the verb is another important factor involved in determining overt vs. zero P marking. That is, whereas overt P marking is optional if P is adjacent to the verb with which it occurs, it is almost obligatory if P is removed from the verb-adjacent position. Thus, if the A argument and the P argument in (474a) is transposed to get a PAV order, then overt P marking becomes obligatory (i.e. kodomo⸗ga kantyoo((⸗oba)) mitoru → kantyoo⸗oba kodomo⸗ga mitoru).

11.2.4 Double Subject Construction

In the Double Subject Construction (DSC), the predicate being either adjectival (verbal-adjectival or nominal-adjectival) or nominal, takes two nominative arguments, what we call the outer subject and inner subject, with the outer subject being topic-marked.

(475)

a.

oraa	zu⸗no	ityaa.
1sg.top	head⸗nom2	hurt.npst

‘I have a headache.’ (lit. ‘As for me, (my) head aches.’) (verbal adjective)

b.

araa	otooto⸗ga	oodoo⸗zyat-ta.
3sg.top	younger.brother⸗nom1	cheeky⸗cop-pst

‘As for him, (his) younger brother was cheeky.’ (nominal adjective)

c.

oraa	oya⸗ga	byooki⸗zyat-ta.
1seq.top	parent⸗nom1	disease⸗cop-pst

‘My parents were ill.’ (lit. ‘As for me, (my) parents had disease.’) (nominal)

The two arguments of the DSC are in a possessive-possessum relationship. DSC is thus a kind of possessor-raising construction whereby the possessor of an NP is raised to the sentential topic.

11.2.5 Transitive Adjectival Construction

Shiiba has a transitive adjectival construction where the adjectival predicate (either a verbal or nominal adjective) takes the experiencer and stimulus (476a, 476b). The two arguments are canonically nominative-marked as in the DSC (§ 11.2.4), but the two arguments in transitive adjectival constructions are not in a possessive-possessum relationship but are independently required by the predicate. A few predicates like ozyee ‘be scared’, sukan ‘dislike’, kitii ‘be distressed’, which commonly depict a stimulus causing a negative feeling in the experiencer, may trigger non-canonical dative marking on the part of the stimulus (see Matsuoka et al. 2019 for details).

(476)

a.

oraa	hanako⸗ga	suki⸗wai.
1sg.top	Hanako⸗nom1	like⸗sfp

‘I like Hanako.’

b.

oraa	hanako{⸗ga/⸗ni}	ozyee⸗wai.
1sg.top	Hanako{⸗nom1/⸗dat}	fear.npst⸗sfp

‘I fear Hanako.’

11.4.1 Causative

The causative suffix -sase- introduces a new participant, the causer, which is coded as A, making the whole sentence transitive (477a) or ditransitive (477b). The original S/A, which is now the causee, is marked with dative case ⸗ni as an E argument (§ 8.2).

(477)

a.

ziroo⸗ga	taroo⸗ni	oyog-ase-ta.
Jiro⸗nom1(Causer)	Taro⸗dat(Causee)	swim-caus-pst

‘Jiro made Taro swim.’ (Causative: transitive)

b.

ziroo⸗ga	taroo⸗ni	tubo⸗oba	waraseta.
Jiro⸗nom1(Causer)	Taro⸗dat(Causee)	pot⸗acc	break-caus-pst

‘Jiro made Taro break the pot.’ (Causative: ditransitive)

The same construction is used for both what is called the ‘make’ causative and the ‘let’ causative. In (477a), for example, one can add an adverb like muriyari ‘against his will’ (make-causative) or sukinasiko ‘as much as he wants’ (let-causative) without any morphosyntactic change.

11.4.2 Passive

The passive-potential suffix -rarE- is used for passivization and for encoding potential expressions (see § 11.5 for potentials). The suffix derives a V-final stem, with the underspecified /E/ being /e/ or /u/. In passivization, the suffix backgrounds the agent, demoting it to an E argument (§ 8.2) marked with dative case ⸗ni or ablative case ⸗kara, while it promotes the original P to S. The demoted agent may often be left unstated.

(478)

ziroo⸗no	tubo⸗no	taroo{⸗ni/⸗kara}	war-are-ta.
Jiro⸗gen2	pot⸗nom2	Taro{⸗dat/⸗abl}	break-pass-pst

‘Jiro’s pot was broken by Taro.’

Shiiba has another passive construction where the possessor of the original P is raised to S (479) (cf. (478)). (479) is intransitive in the sense that the nominative argument (ziroo) is not the agent of the action denoted by the verb war- ‘break’, but the possessor of the patient (tubo ‘pot’), who is affected by the event of Taro breaking his pot. The agent is now demoted and dative-marked. It is unclear whether it can also be marked by ablative ⸗kara as in the case of the direct passive (478).

(479)

ziroo⸗ga	taroo⸗ni	tubo⸗oba	war-are-ta.
Jiro⸗nom1	Taro⸗dat	pot⸗acc	break-pass-pst

‘Jiro had his pot broken by Taro.’ (Possessor passive)

11.6.1 Tense

Tense opposition is between past and non-past. The non-past tense covers present reference (e.g. tigau (tigaw-[ru]) ‘be different’), atemporal statements (e.g. higemusi⸗ni sasareta⸗naryaa te⸗no tadaruru (tadar{e/u-[ru]) ‘If you get bitten by a caterpillar, your hand gets inflamed.’), and future reference (e.g. tegami⸗wa ato⸗de yomu (yom-[ru]) ‘(I) will read the letter later.’).

11.6.2 Aspect

The major aspectual opposition is between the perfective and the imperfective. The perfective aspect, which construes the situation as an independent whole, is coded simply using a tense-marked finite verb or by the sequential converbal form (482). The imperfective aspects construe the situation as having its internal structure, focusing on its inceptive, durative and completive phases with the former two coded with a syntactic compound with the progressive root or- and the last one with the completive suffix -tor-.¹⁵

(482)

imooto⸗ga	kimono⸗oba	aroo-te	hwee-ta.
younger.sister⸗nom1	clothes⸗acc	wash-seq	dry-pst

‘My younger sister washed the clothes and dried them.’ (Perfective)

The progressive or- construes the situation as being in a durative phase, i.e. a phase in which the event is ongoing (483a). In the cases of change-of-state verbs (achievement and accomplishment verbs), the durative phase means the event has not yet exceeded the point of change-of-state, gradually or abruptly approaching the end point (483b). In the past tense, the progressive may designate habituality, which can be analyzed as an unbounded (ongoing) repetition of a certain event (483c).

(483)

a.

musi⸗no	nak-i-or-u.
insect⸗nom2	make.noise-prog-npst

‘There’s an insect making noise.’

b.

musi⸗no	sin-i-or-u.
insect⸗nom2	die-thm-prog-npst

‘There’s an insect dying.’

c.

mukasyaa	koko-hen⸗wa	musi⸗no	nak-i-ot-ta.
in.old.days.top	this.place-around⸗top	insect-nom2	make.noise-prog-pst

‘In the past, insects used to make noise around here.’

The completive -tor- construes the situation as having exceeded the point of change of state described by the verb and having entered a stable, resulting phase (i.e. a perfect phase). If applied to achievement verbs, it always designates a phase after the end point of the change-of-state (484a), giving rise to a clear aspectual contrast between the progressive and the completive (compare (483a) and (484a)). If applied to accomplishment verbs, it may designate either perfect or durative, as illustrated in (484b). This is due to the fact that accomplishment verbs refer to both a beginning point and an end point, and the completive may construe the situation as having exceeded either the beginning point (in which case the situation is construed as having entered the durative phase) or the end point (in which case the situation is construed as having entered the perfect phase). In the case of activity verbs (484c), which refer to only a beginning point but lack an end point, the completive designates the durative phase of the event.

(484)

a.

musi⸗no	sin-dor-u.
insect⸗nom2	die-cpl-npst

‘There’s an insect dead.’

b.

{mada/moo}	mi-tor-u
{still/already}	watch-cpl-npst

‘(someone) is still watching (durative phase)/has already watched (perfect phase).’

c.

musi⸗no	nyaa-tor-u.
insect⸗nom2	make.noise-cpl-npst

‘There’s an insect making noise.’

11.6.3 Modality

Modalities are expressed by inflection (mood), modal particles or by complex predicates. The moods are the indicative, intentional and imperative. The intentional mood covers both speaker’s intention (e.g. ore⸗ga toroo ‘I’ll take’) and locutors’ intention (e.g. (tyende) toroo⸗ya ‘let’s take (together)’). Conditional converbal inflection designates a deontic-modal meaning, as in tor-a-nyaa (take-thm-neg.cond) ‘(you) need to take’ (lit. ‘if (you) don’t take’).

As mentioned in § 10, the present authors have identified the following modal particles: ⸗huu (inferential; Kato 2017), ⸗tyuu (hearsay), ⸗gotaru (similative; ‘seems like’), ⸗doo (conjectural). We have not collected enough data to describe the function of each morpheme, though it is possible to state at this stage that all of these forms pertain to epistemic as opposed to deontic modality. The bulk of deontic modalities are expressed through mood inflection (intentional, imperative) and various kinds of complex predication (485b).

(485)

a.

hito⸗no	nak-i-or-u{⸗huu/⸗tyuu}⸗wai.
person⸗nom2	cry-thm-prog-npst{⸗infr/⸗hs}⸗sfp

‘(I heard that) there’s a man crying.’ (⸗tyuu: hearsay)

‘(I have certain evidence (e.g. sound) that) there’s a man crying.’ (⸗huu: inferential)

b.

kore	[kat-ta⸗ga	ee]⸗ga.
this	buy-pst⸗nom1	good.npst⸗sfp

‘It’s better for you to buy this.’ (Deontic: suggestion)