Kharia language

Not to be confused with Kharia Thar language or Kharia, Jalpaiguri.

Kharia
खड़िया, ଖଡ଼ିଆ
Region	India (Jharkhand, Chhattisgarh, Odisha).
Ethnicity	Kharia
Native speakers	297,614; 69% of ethnic population (2011 census)
Language family	Austroasiatic Munda South Kharia
Writing system	Devanagari, Odia, Latin
Official status
Official language in	India Jharkhand (additional)
Language codes
ISO 639-3	khr
Glottolog	khar1287
ELP	Kharia

A multilingual person speaking Sadri, Kharia, and Sambalpuri language, recorded in China.

The Kharia language (Kharia pronunciation: [kʰaɽija] or [kʰeɽija]) is a Munda language of the Austroasiatic language family, that is primarily spoken by the Kharia people of eastern India.

History

The first systematic description of the Kharia language is Banerjee (1894)'s Kharia grammar, followed by Tea Districts Labour Association (1929) and Floor et al. (1934), which resulted in a Kharia-English Dictionary. An ethnological study on the tribe was published in 1937 by Roy & Roy.

The first major academic approach to Kharia were taken by linguist Heinz-Jürgen Pinnow in the 1950s and 1960s with studies published in both German and English. Other works include Biligiri (1965)'s full study and lexicon; Mahapatra (1976) on Kharia and Juang verbs, Malhotra (1982) Ph.D. dissertation attempting a comprehensive grammar of Kharia; Abbi (1993; 1997) on language change and contact; Rehberg (2003) on Kharia phonology (in German).

Classification

Kharia belongs to the Kharia–Juang branch of the Munda language family. Its closest extant relative is the Juang language, but the relationship between Kharia and Juang is remote.

Kharia is in contact with Sadri (the local lingua franca), Mundari, Kurukh, Hindi, and Odia (in Odisha).

Distribution

Kharia speakers are located in the following districts of India.

• Jharkhand
  • Simdega district
  • Gumla district
• Odisha
  • Sundargarh district
  • Jharsuguda district

Kharia bidaai song

Phonology

Kharia consonants

		Labial	Dental/ Alveolar	Retroflex	Post-alv./ Palatal	Velar	Glottal
Nasal		m	n	(ɳ)	ɲ	ŋ
Stop/ Affricate	voiceless	p	t̪	ʈ	c	k	(ʔ)
	aspirated		t̪ʰ	ʈʰ	cʰ	kʰ
	voiced	b	d̪	ɖ	ɟ	ɡ
	breathy		d̪ʱ	ɖʱ	ɟʱ	ɡʱ
	glottalised	ˀb		ˀɖ	ˀɟ
Fricative	voiceless	f	s
	voiced	v					ɦ
Approximant		w	l		j
Tap	unaspirated		ɾ	(ɽ)
	aspirated			(ɽʱ)

• [ɽ, ɽʱ] are only marginally phonemic and are normally intervocalic allophones of /ɖ, ɖʱ/.
• /f/ came from earlier */pʰ/ and can also be pronounced among some speakers as an affricate [p͡f], which was most likely an intermediate stage between the /pʰ > f/ shift.
• /v/ came from earlier */bʱ/ and can also be pronounced among some speakers as an affricate [b͡v], which was most likely an intermediate stage between the /bʱ > v/ shift.
• /c, cʰ, ɟ, ɟʱ/ are often realized as affricate sounds [t͡ʃ, t͡ʃʰ, d͡ʒ, d͡ʒʱ], especially in loanwords.
• [ʔ] is an allophone of /ɡ/ when in coda position.

Kharia vowels

	Front	Central	Back
Close	i		u
Mid	e	(ə)	o
Open		a
Diphthong	/ae̯, ao̯, ou̯, oe̯, ui̯/

• /i, e, o, u/ have lax allophones of [ɪ, ɛ, ɔ, ʊ].
• /a/ can have allophones of [ɑ, ä, ə, ʌ].

Gemination only occur in morpheme boundaries of words. Consonant length can be phonemic. E.g. /oton=na/ realized as [ɔtɔnːɑ] (press=INF). /ʔ/, /s/, and /h/ may not be geminated.

Morphosyntax

Parts of speech in Kharia

Issues with parts of speech in Kharia

In most languages, there are basic rules and distinctions among lexical categories such as nouns and verbs. Nouns, the central nucleus of noun phrases, primarily serve referential functions, whereas verbs function predicatively. In some languages, however, lexical items are not inherently specified for such functions. When a lexeme can flexibly function referentially, predicatively, or attributively without derivation, and fails to be lexically specified morphosyntactically, it is known as precategoriality. Languages in which the majority of the lexicon or the majority of complex syntactic are precategorial are therefore called precategorial languages. Unlike English, where derivational processes are highly productive (such as the verbalizer -ify), and predication typically requires copulae (i.e., to be) or auxiliary verb constructions, precategorial languages do not rely on these mechanisms, as underived lexemes can directly assume referential, predicative, or attributive functions without category-changing morphology.

There is little if any structural evidence in Kharia to support the existence of a noun–verb–adjective distinction in (Simdega) Kharia. It seems that the Kharia lemmas' functional distribution is organized by syntactic configuration rather than inherent lexical property. The information carried by the lemma does not determine its inherent syntactical and lexical category. Virtually all roots, from basic monosyllabic lexemes, derived words, to complex phrases, regardless of potential meanings and functions and discourse choices, can all freely serve as semantic heads in either a predicate or its complements (arguments and modifiers) with compositional results. There are no morphological distinctions between these uses.

"Generic nouns":

lebu

man

ɖel=ki

come=MID.PST

lebu ɖel=ki

man come=MID.PST

'The/A man came.'

Bhagwan

God

lebu=ki

man=MID.PST

ro

and

ɖel=ki

come=MID.PST

Bhagwan lebu=ki ro ɖel=ki

God man=MID.PST and come=MID.PST

'God became man (Jesus) and came (to earth).'

ho=kaɽ=te

that=SG.HUM=OBL

iɲ=aʔ

I=GEN

natgoth=oˀj

family.relationship=ACT.PST.1SG

ho=kaɽ=te iɲ=aʔ natgoth=oˀj

that=SG.HUM=OBL I=GEN family.relationship=ACT.PST.1SG

'I accepted him/her as a member of my family'

Wh-interrogatives:

gam=e=pe

say=ACT.IRR=2PL

i=ye=niŋ?

what=ACT.IRR=1PL.INCL

gam=e=pe i=ye=niŋ?

say=ACT.IRR=2PL what=ACT.IRR=1PL.INCL

'Tell me, what shall we (do)?'

Indefinites:

jahã

INDEF

guɽuʔ=may[...]

OPT=3PL

jahã guɽuʔ=may[...]

INDEF OPT=3PL

'Let them (do) whatever[...]'

Quantifiers:

jhaɽi

all

daru

tree

biru

mountain

oɖoʔ

other

moɲ

one

goɖ=ki=may

CTEL=MID.PST=3PL

jhaɽi daru biru oɖoʔ moɲ goɖ=ki=may

all tree mountain other one CTEL=MID.PST=3PL

'All trees and mountains and everything else(other) became one.'

Property modifiers:

konon=na=ba

small=MID.IRR=2SG.HON

konon=na=ba

small=MID.IRR=2SG.HON

'Become small/Make yourself small!'

Proper names:

aʔghrom=ki

Aghrom=MID.PST

aʔghrom=ki

Aghrom=MID.PST

'It came to be called(became) Aghrom' (Literally 'It Aghromed.')

aʔghrom=oʔ

Aghrom=ACT.PST

aʔghrom=oʔ

Aghrom=ACT.PST

'Someone named it Aghrom' (Literally 'Someone Aghromed it.')

Status and roles:

ho=kaɽ

that=SG.HUM

ayo=ki

mother=MID.PST

ho=kaɽ ayo=ki

that=SG.HUM mother=MID.PST

'She became mother.' (Literally 'That person mothered.')

ho=kaɽ=te

that=SG.HUM=OBL

ayo=yoˀj

'mother=ACT.PST.1SG

ho=kaɽ=te ayo=yoˀj

that=SG.HUM=OBL 'mother=ACT.PST.1SG

'I accepted her as mother.' (Literally 'I mothered that person.')

Deictic "adverbials" and proforms:

musa

today

iɖaʔ=ki

yesterday=MID.PST

musa iɖaʔ=ki

today yesterday=MID.PST

'Today became yesterday' (Literally 'Today yesterday-ed.')

ho

that

rochoˀb=ki=ɲ

side=MID.PST=1SG

ho rochoˀb=ki=ɲ

that side=MID.PST=1SG

'I moved to that side.' (Literally 'I that-side-ed.')

am=ga

You=FOC

am=na=m

You=MID.IRR=2SG

am=ga am=na=m

You=FOC You=MID.IRR=2SG

'Thou will be thou (Literally You beyou will).'

"NP" with genitive:

am

You

ho=kaɽ=te

3=SG.HUM=OBL

am=aʔ=yoˀb

you=GEN=ACT.PST.2SG

am ho=kaɽ=te am=aʔ=yoˀb

You 3=SG.HUM=OBL you=GEN=ACT.PST.2SG

'You adopted him/her (Literally You made him/her yours)'

Genitive attribute in "NP":

Bharat=yaʔ

India=GEN

lebu=ki

person=PL

bides=aʔ

aboard=GEN

lebu=ki=yaʔ

person=PL=GEN

rupraŋ=ki=may

appearance=MID.PST=3PL

Bharat=yaʔ lebu=ki bides=aʔ lebu=ki=yaʔ rupraŋ=ki=may

India=GEN person=PL aboard=GEN person=PL=GEN appearance=MID.PST=3PL

'The Indians took on the appearance of foreigners (e.g. by living abroad so long).'

What listed above are quite deviated far away from the generally conceivable concepts of "noun" or "nominal", which are usually known as a lexical or syntactical category prototypically associated with the semantic category of thing, and "verb" or "predicative", which are usually known as a lexical or syntactical category prototypically associated with the semantic category of action. Of course, one from general prescriptivist linguistics could assume that (Simdega) Kharia's case is similar to English's zero-derivation/conversion strategy, which shifts word categories through verbalization and deverbalization without morphological change, and thus asserts that Kharia must have lexical categories in the sense of mainstream Indo-European-based linguistic frameworks as well. However, as examples from the corpus show that even complex NP-like structures like "appearance of aboard people" and proper names like Aghrom can appear as content heads of predicative phrases with compositional semantic results, suggesting that Kharia does not "use nouns as verbs" or "use verbs as nouns", or "there is a hidden verb"/"hidden copula", but its lexicon neither fits into the mainstream pedagogical understanding of lexical classes. Rather, the overarching prescriptivist view that all languages share the same underlying structures has never provided a satisfactory and exhaustive structural explanation for Kharia. Furthermore, the next section provides additional evidence for why derivations in Kharia do not create new stable word classes or assign different lexical category to derived forms.

Derivations

Lexical juxtapositions

In Kharia, lexical juxtapositions are constructions formed by arranging two (or more) contentive morphemes side by side without morphological bonding, but result in a single semantic unit. Their meanings are often compositional or loosely associative, e.g., combinations such as "forest + scrubland" meaning 'wilderness' or "field + farm" creating 'farmland', depending on interpretation. Although collectively showing a subtle semantic outcome, the two contentive morphemes do not constitute a single prosodic and morphological word. Furthermore, the juxtaposed contentive morphemes can be easily separated by a clitic or another word.

Lexical juxtapositions	Word-by-word meaning	New meaning
ompay kirom	big.river + river	'rivers'
kheti baɽi	field + farm	'farms'
kinir jhaŋkoy	forest + scrubland	'forest, wilderness'
goˀjlo[ʔ] aŋkel	field + land	'cultivated land'
baʔ rumkhuˀb	paddy + rice	'grain'
paisa ɖhebuwa	money (Indo-Aryan) + money (Kharia)	'wealth'
usloʔ tirib	earth + cloud	'universe'
tiʔ kaʈa	hand + leg	'body parts'
kulam kulamɖay	brother + sister	'people'
oreˀj koŋtaŋ	bull + cow	'cattle'

True compounds

Due to bimoraic constraint, Kharia, like most Munda languages, requires phonological words to be at least bisyllabic, with very few exceptions. Thus, virtually all standalone contentive morphemes consist of at least two syllables. True compounds (like the one listed below) in Kharia are actual phonological words with single coherent prosodic profiles. They are typically formed from two concatenated morphemes, one of which is the monosyllabic lexical core of a contentive morpheme. This lexical core is phonologically dependent and cannot occur in free isolation.

Kharia compound
	Compounds	Morpheme-by-morpheme meaning	New meaning
Structurally considered compounds	baʔ-loʔ	paddy-land (< usloʔ "land")	"paddy"
	jaʔ-loʔ	grass-land	"meadow"
	dhuɽi-loʔ	dust-land	"desert"
	kunɽu[ʔ]-siŋ	child-fowl (< siŋkoy "chicken")	"chick"
	si-loʔ	plow-land	"plow"
	paʔ-sor	break-stone (< soreŋ "stone")	"grinding slab"
	muʔ-ɖaʔ	emerge-water	"source, origin(ate)"
	soʔ-lui	tie-hair (< ului "hair")	"tie up a girl's hair; girl's hair"
	ur-boʔ	shake-head (< bokoˀb "head")	"shake the head"
	iˀj-thaŋ	shit-cow	"(cow) dung"
Semantically unpredictable compounds	diyom-jaŋ	lamp-bone	"knee bone" (lit. "lamp-shaped bone")
	goŋɽiŋ-oʔ	cook-house	"kitchen"
	kaɖoŋ-moˀɖ	fish-eye	"name of a tree (whose leaves are shaped like fish eyes)"
	borol-ɖaʔ	live-water	"fresh water"

Pseudo-compounds and "noun incorporation"

In Kharia, pseudo-compounding (which may be referred to as "incorporation") of semantic bases and adjuncts has been fossilized and mostly limited to certain stems and to lexicalized (non-productive) forms. Polysyllabic contentive words are reduced to their final syllable(s), while no phonological adjustments occur on monosyllabic items. The incorporated compounds may result in idiosyncratic semantics.

There are two main types of these constructions: historical and pseudo-compounds. Neither of these types is immune to precategoriality as they remain contentive and fully opaque to syntax, although they tend to occur in predicative function. The internal structure of historical compounds may reveal an earlier stage of Kharia, when the language may still have retained an S-V-O structure and had distinct lexical categories of verbs and nouns inherited from its ancestral Proto-Austroasiatic predecessor.

The "pseudo-compounds" and semi-productive incorporation contentives (4, 5 and 6) originated from a later stage of the language. Additionally, there are few forms (7) that were derived from incorporation whose "verbal" component no longer exists as a contentive morphemes.

Historical:

1. guˀj 'wash' + tiʔ 'hand' = guʔthe, guʔʈhe 'wash hand'

2. boŋ 'finish' + sor (< soreŋ ('stone')) = boŋsor 'petrify'

3. muʔ 'emerge, come out' + siŋ (< siŋkom 'star') = muʔsiŋ '(the sun) to rise'

Pseudo-compounds:

4. ajoˀɖ 'dry' + ɖaʔ 'water' = ajoˀɖ ɖaʔ 'dry up'

ɖaʔ

water

ajoˀɖ

dry

ɖaʔ=ki

water=MID.PST

ɖaʔ ajoˀɖ ɖaʔ=ki

water dry water=MID.PST

'The water dried up.'

5. betoˀɖ 'become hungry' + ɖaʔ 'water' = betoˀɖ ɖaʔ 'become thirsty'

Semi-productive incorporation:

6. ulaʔ 'letter' + likha 'write' = ulaʔ likha 'write letter'

kol

RECP

ulaʔ

letter

likha=ki=kiyar

write=MID.PST=DU

kol ulaʔ likha=ki=kiyar

RECP letter write=MID.PST=DU

'They wrote each other letters.' (accepted by some speakers. Preferred form ulaʔ kol likha=ki=kiyar).

Incorporation with lost "verbal" component:

7. kul-ɖaʔ 'get/have a fever' < *kul ? ɖaʔ 'water'

From a historical linguistic perspective, pseudo-compounds were likely derived later, imitative of historic compounds, thereby mimicking their internal syntactic structure. The semi-incorporations were likely created more recently, as evidenced by their OV order and limited usage (accompanied with the reciprocal marker kol). Still, the pseudo-compounds and semi-incorporations are merely juxtaposed contentive morphemes and do not share the properties of a compound, that is, a single structurally integrated phonological and syntactic unit. A floating clitic can be freely inserted between two components of pseudo-compounds. Such frequent quirks invalidate their phono-morphological status as compounds.

ɖaʔ

water

ajoˀɖ=ga

dry=FOC

ɖaʔ=ki

water=MID.PST

ɖaʔ ajoˀɖ=ga ɖaʔ=ki

water dry=FOC water=MID.PST

'The water dried up.'

urum=ga

warm=FOC

ɖaʔ=ki

water=MID.PST

urum=ga ɖaʔ=ki

warm=FOC water=MID.PST

'S/he sweated.'

On the other hand, enclitic markers cannot be inserted between two elements in historical compounds. The elements of these forms are tightly bonded syntactically and thus are justifiably considered as single monolithic phonological units.

boŋ-sor=ki

finish-stone=MID.PST

boŋ-sor=ki

finish-stone=MID.PST

'It petrified.'

ol-ɖay=oʔ

take-woman=ACT.PST

ol-ɖay=oʔ

take-woman=ACT.PST

'He married.'

Overt prefixing and infixing derivation

Austroasiatic languages are characterized by rich and historically productive systems of derivation, especially through prefixation and infixation. In many branches of the family, particularly in Mainland Southeast Asia, these derivational processes are associated with lexical categorization, producing forms that fall into fixed syntactic categories of nouns, verbs, or other parts of speech. Affixation in those languages often involves nominalization, verbalization, or other category-changing operations.

In Kharia, although cognate derivational morphology is formally present, its functional outcome is significantly different. It also hardly resembles that of the Mon-Khmer derivation. Lexemes created through prefixation or infixation, e.g., with the nasal infix -n- or causative prefixes -(o)(ʔ)b- produce a new lexical concept from another, somewhat related in semantics but unpredictable. However, these affixes do not assign nor change the syntactic category of the derived lexemes. Derived lexemes may denote instruments, locations, events, or causative relations, but they remain distributionally flexible and may occur as semantic heads in referential, predicative, or attributive functions.

Underived contentive morphemes	Meaning	Derived contentive morphemes	Meaning
joʔ	'sweep'	jo-n-oʔ	'broom'
bel	'spread out'	be-n-el	'bedding'
muʔsiŋ	'rise'	mu-n-uʔsiŋ	'east'
jin	'touch'	ji-n-ib	'touch'
bui	'raise, keep'	bu-n-ui	'pig'
rab	'bury'	ra-n-ab	'burial ground'
roʔ	'drop, spill'	ro-n-oʔ	'droping/spilling'
tol	'tie'	to-n-ol	'string, toe face'
*sel	'pray'	se-n-el	'prayer'
gil	'beat'	gi-n-il	'beating'
uʔfe	'three'	uʔ-n-fe	'three each'

Due to pragmatic factors, it is nearly impossible to observe derived morphemes such as bunui 'pig' functioning as predicates in everyday discourse. However, Simdega Kharia speakers do not perceive using derived contentive morphemes in flexible constructions as illgrammatical, showing that the flexible construction is still structurally legitimate. Thus, bunui may have the predicative meaning "to become a pig" in middle voice, and "to make someone/something a pig" in active voice, while in referential function, it simply means "a/the pig" (cf. ginil 'beating, beat (habitually)). In summary, except for a small, closed class of proforms and deictics of time and place, there is no syntactic difference between the underived and derived forms as they both remain precategorial. Thus, the affixes in Kharia do not display any characteristics of "nominalizers" or "adjectivizers" nor "verbalizers.".

Echo-words

Echo words in Kharia are formed through partial reduplication, in which the second element mirrors the phonological structure of the first with systematic variation. These constructions typically express something resembles English "etc.". E.g. main marjad 'honour, etc.', kado walo 'mud, etc.'

Echo-formation does not assign lexical category status. The resulting form remains a contentive unit and may serve as the semantic head of predicative construction, either a partially finite or fully finite. The echo-word may be separated from the lexical head by variously morphemes when it is used as an imperfective converb, evidently suggesting that echo formations are not compounds.

musniŋ

one.day

ho=ki

that=PL

jhaɽi

all

kuruʔ

child

hakon=ki=te

ECHO=PL=OBL

somtay=oʔ=ki

collect=ACT.PST=PL

ro

and

kuda

hurry

no

CMPL

kudi

ECHO

raylogaɽh

Railogarh

tay

ABL

yar

flee

kan=oʔ=ki

CONT=ACT.PST=PL

musniŋ ho=ki jhaɽi kuruʔ hakon=ki=te somtay=oʔ=ki ro kuda no kudi raylogaɽh tay yar kan=oʔ=ki

one.day that=PL all child ECHO=PL=OBL collect=ACT.PST=PL and hurry CMPL ECHO Railogarh ABL flee CONT=ACT.PST=PL

'One day they gathered up all the children and fled quickly (hurrying) from Railogarh.'

Defining Kharia lexical categories

To summarize, the Kharia lexemes show equally distributional and semantic regularity, stable and compositional, across syntactic functions—under parametrical and structural analysis— and no consistent set of ontological, semantic, syntactic, or discourse-pragmatic criteria supports evidence that any verb-noun distinction exists. Henceforth, linguist Peterson (2011) concludes that traditional lexical categories such as verbs and nouns are formally absent in Kharia. Instead, according to his analysis, the Kharia lexicon is divided into two principal classes:

The contentive morphemes form an open class of semantically full elements that carry lexical meaning. They are inherently underspecified for grammatical category and can occur as the SEMANTIC/CONTENT HEAD in referential, predicative, or attributive function. The syntactic and morphological environment determines their role within the clause. When a contentive morpheme is marked with TAM/PERSON enclitics, it functions predicatively. Its combination with case and number enclitics produces a referential interpretation.

The functional morphemes and proforms/deictics form a closed class of functional elements that mark grammatical functions and also signal the syntagma's syntactic role. They include enclitic markers for case, number, tense–aspect–mood, person, honorific status, and various discourse functions. Functional morphemes cannot occur independently and do not contribute lexical meaning. They attach phonologically and syntactically to the contentive morphemes or clauses, forming the FUNCTION HEADS of Case-Syntagmas and TAM/PERSON-Syntagmas. Their attachments determine the syntactic and morphological properties of the clause, defining grammatical relations and predicate structure.

Differences in perceptions over parts of speech in Kharia dialects

• 
  Location of Simdega district of Jharkhand. Speakers of the Simdega dialect generally accept flexible constructions.
• 
  Location of Gumla district. Speakers of the Gumla Kharia dialect tend not to tolerate flexible constructions.

The Kharia lexicon shows dialect-specific variation in how speakers conceptualize lexical organization, in particular regarding distinction between lexical categories. The investigation conducted by Peterson found that, although the two main varieties of Kharia share the same morphosyntactic structures, the degree of lexical flexibility varies between them.

In the Simdega dialect, lexical items fall into an open class of contentive morphemes and a closed class of proforms/deictics and functional markers. No independent noun-verb distinction is evident, and categorization is determined by syntactic configurations. Any contentive morpheme, including reduplicated masdars, may serve freely in referential, predicative, or attributive roles.

The Gumla dialect presents clearer evidence for a noun-verb distinction. Speakers generally do not accept predicative uses of lexical items referring to physical or animate entities without a light verb. They avoid complex predicates built from nominal expressions. Still, many contentives remain lexically underspecified, and both dialects share the same structural units—Case-Syntagmas and TAM/person-Syntagmas. Dialectal differences may emerge primarily in the acceptability and complexity of potential semantic heads.

Another factor concerning legal acceptability is age. Documentation interviews found that, among adult male and younger Simdega Kharia speakers, many of whom are school students, the acceptability rate is always 100%. Older adults (especially women) and elderly speakers who more strictly adhere to the "pure" or "classical" Kharia (Sudh Kharia) tend to reject those constructions. However, Classical Kharia is a variety of traditional written Kharia found in old literature, it is not spoken, and it is highly influenced by Hindi and Sadri. Moreover, earlier texts collected by Pinnow (1965) and Kerketta (1990) show that these disputed constructions did in fact occur in spontaneous and literary Kharia, suggesting that the flexible interpretation of Kharia is structurally legitimate for both the spoken and literary forms of the language. Sociolinguistic factors, prescriptivist attitudes, and language contact may reinforce the idea that views flexible constructions as "inelegant" and thus many adult speakers tend to avoid.

The core Kharia-speaking area of the Simdega district overlaps with the North Munda (Mundari) zone. The language has been in prolonged contact with Mundari, having undergone areal convergence with North Munda and Sadri. Today, Kharia exhibits significant influence from Mundari morphosyntax and lexicon, especially regarding the parts of speech, with respect to precategoriality. North Munda, especially Kherwarian lects of Jharkhand, typically show virtually absence of evidence for lexical categories such as nouns, verbs, adjectives. South Munda languages, however, at least tend to distinguish nouns and verbs, except Kharia, so the only South Munda language in Jharkhand. Conversely, the northern and southern peripheral Kharia dialects display lesser lexical flexibility. In terms of this kind of extreme lexical flexibility, Simdega Kharia is undoubtedly unique within its subgroup.

Case-Syntagmas ("Noun phrases")

The Case-Syntagma is a morphosyntactic structure consists of a semantic head and a functional head which is composed of case and number enclitics. The semantic head can be a single contentive morpheme or a more complex phrasal unit, including genitive-marked constructions. The function head includes case, possessive, and number markers attached to the rightmost element of the syntagma, governing the Case-Syntagma's grammatical information independently of the content head's lexical meanings.

The Case-Syntagma covers functions typically associated with arguments, possessors, and modifiers. It contains reference and nominal relations, but differs fundamentally in that it is not anchored in a lexically defined "noun" category. Its nominal behavior is solely provided by the functional morpheme clitics. It is broadly comparable to a noun phrase in English.

A schematic overview of the Case-Syntagma Structure in Kharia:

GEN-ATTR DET DEM QUANT=CLF GEN-ATTR DET Lexeme(s)=POSS=NUM/HON=GEN=CASE

Pronouns

		Singular	Dual	Plural
1st person	exclusive	iɲ/iŋ (less common)	iɲjar	ele
	inclusive		anaŋ	aniŋ
2nd person		am	ambar	ampe
3rd person	Anaphoric	aɖi	aɖ(i)kiyar	aɖiki
	unmarked	hokaɽ, hojeʔ ukaɽ, ujeʔ hinkaɽ, hinjeʔ hankaɽ, hanjeʔ	hokiyar ukiyar hinkiyar hankiyar	hoki uki hinki hanki

Case

Kharia Case-Syntagma has three cases:

• Nominative: unmarked
• Oblique: =te
• Genitive: =aʔ

The oblique case marker =te denotes a distinction between core and peripheral arguments. Participants marked with =te are interpreted as structurally and semantically external, in contrast to the nominative unmarked arguments. Thus, a Case-Syntagma marked with an oblique case marker is effectively blocked from forming part of the content head of a TAM/Person-Syntagma.

sahar=te=ki=ɲ

city=OBL=MID.PST=1SG

sahar=te=ki=ɲ

city=OBL=MID.PST=1SG

*Not applicable

The genitive marker =aʔ in Kharia does not mark the relationship of a Case-Syntagma to other syntagmas, such as a TAM/Person-Syntagma or another Case-Syntagma. Instead, it signals that the marked semantic base may function as part of a larger semantic head or that a certain postposition selects a semantic head in the genitive.

oʔ=yaʔ

house=GEN

teloŋ

roof

oʔ=yaʔ teloŋ

house=GEN roof

'the roof of the house'

og=aʔ

house=GEN

bahare

outside

og=aʔ bahare

house=GEN outside

'outside of the house'

Gender

Grammatical gender is not a morphosyntactical feature of Kharia, but the language has independent words to identify whether a male or female of a lexical word is intended. E.g. kokro siŋkoy 'rooster' and kitur siŋkoy 'hen'.

Person

Inalienable content heads are cross-referenced with possessive markers showed in the table below.

		Singular	Dual	Plural
1st person	exclusive	=ɲ/iɲ/(i)ŋ	=jar	=le
	inclusive		=naŋ	=niŋ
2nd person		=nom	=bar	=pe
3rd person		=ɖom	=ɖom=kiyar	=ɖom=ki

Interrogatives

	Interrogatives
ata	'what?, which?'
atu	'where?'
ber, behar	'who?'
i	'what?'
ina	'why?'
a-	Question marker

Numerals

Kharia has two numeral systems. The one native to Kharia is no longer in common productive use, therefore having great disparities and disagreements. The other, which was borrowed from Sadri, is used in daily life.

	native numerals	borrowed from Sadri
0		sun
1	moɲ (NHUM), muɖu (HUM)	ek
2	ubar	dui
3	ufeʔ	tin
4	ifonʔ, tham	cair, ceir
5	moloy, thum	pãc
6	tibru, tibhru, ʈibru	chaw
7	ghul, tham, thom, thoŋ	sat
8	tham, thom, thomsiŋ, ghul	aʈh
9	thomsiŋ, tomsiŋ, ghal, ghul	naw, nãw
10	gʰol	das
100	moloy ekɽi	say, saw, sos
1000		hajar

The Sadri derived numerals often go with numeral classifiers. Classifiers occur very seldom with native numerals, at least by modern speakers, perhaps due to the unfamiliarity of the modern speakers with the Kharia numerals.

Kharia classifiers

Classifier	Note
cepuʔ	'handful'
gaɽ	possibly derived from kaɽ 'human'
go	Not attested by Peterson
jan/jhan	used with human entities
mene	multifunctional
=o	general classifier, clitic
ʈaŋ/ʈhaŋ/ʈhoŋ	similar to ʈho
ʈho	generally non-human entities

Except mene and gaɽ, the mentioned classifiers of Kharia are well in line with the classifier systems of neighboring Eastern Indo-Aryan languages (notably Sardi), with cognate forms suggesting they were borrowed from Indo-Aryan. The use of classifiers in Kharia is optional, however. They often stand with cardinal numerals, and rarely, with quantifiers.

For the case of mene, Peterson describes: "mene is the most elusive of the classifiers, and in fact, its status as a classifier is questionable." Mene can occur with not only different quantifiers but also any type of lexical heads. It may be found occurring with demonstratives, a genitive attributive, or even an adverbial. While its true function is still unknown, Peterson tentatively considers it a classifier.

TAM/Person-Syntagmas ("Verbs")

Kharia lacks lexical categories of nouns and verbs, so any semantic bases/contentive morphemes, and even a whole complex Case-Syntagma structure corresponding to the ordinary "noun phrase" can function as the head in the predicative role within a syntactic phrase. In assembly with other functional morphemes, they form a TAM/Person-Syntagma, the predicative syntactic unit of a sentence. There are two major types of predicates in Kharia: narrative predication, which may resemble the typical predication in most languages, and equational predication (approximately "be" construction). The latter subtype consists of non-inherent qualitative predication and qualitative predication, both of which share the same structure as the narrative predicates except in the present tense. The qualitative predicative marker may also not be present, or "lacking", resulting in what resembles a "nominal sentence" but with subtle semantic differences to examples in which an element translating as 'be' or 'become' is present.

The simplified maximal structure of a non-periphrastic, fully finite narrative TAM/Person-Syntagma in Kharia:

RECP CAUS-Lexeme/CAUS/ v2=PERF=TAM/VOICE=PERS/NUM/HON

The maximal structure of a fully finite, non-negative, non-periphrastic TAM/Person-Syntagma can be described as follows:

RECP CAUS-Lexeme/CAUS/ EXCESS BEN CONT/ITER/SITER PASS/REFL AUTOPOES TEL=PERF=TAM/VOICE=PERS/NUM/HON

That is, the content head of the syntagma can be a contentive morpheme or a complex Case-Syntagma structure, potentially modified by a causative prefix or infix. Following and supporting the semantic head is the functional head, containing a series of explicator verbs or "v2", although their status as phonological words is somewhat unclear with respect to intonation, and many of them do not seem to carry any lexical meanings other than to mark voice, aspects, aktionsart, and orientation to the TAM/Person-Syntagma. Functional morphemes such as TAM and Person/Number/Honorific are marked by enclitics attached to the rightmost element of the function head.

Types of Semantic bases in TAM/Person-Syntagmas

The most common type of semantic head of a finite TAM/Person Syntagma is a simple, non-derived contentive morpheme:

gam=te=ki

talk=ACT.PRES=3PL

gam=te=ki

talk=ACT.PRES=3PL

'They say'

A monosyllabic contentive morpheme can reduplicate to derive a masdar. The masdar TAM/Person-syntagma is used to denote various tenuous forms of TAM-aktionsart, such as remote past, habitual, or an event that took longer than usual. Masdars can only take the middle TAM markers.

lam~lam=ki

RDPL~search=MID.PST

lam~lam=ki

RDPL~search=MID.PST

'S/he looked for [it] for a long time.' (Accepted in Simdega Kharia, rejected in Gumla Kharia)

Subject marking

Similar to Remo, Gutob, Gtaʔ, and recently Juang, Kharia predicate only marks person/number of the subject argument. Distinction between animate and inanimate agents is not so profound in Kharia as they are both marked, although Biligiri (1965) stated that "there is a stronger tendency to observe number agreement with an animate subject than with an inanimate subject."

		singular	dual (HON)	plural
1st person	exclusive	=ɲ(iɲ)/=ŋ(iŋ)	=jar	=le
	inclusive		=naŋ	=niŋ
2nd person		=(e)m	=bar	=pe
3rd person		=Ø	=kiyar	=ki/=may

Tense, aspect, mood

Kharia, like many Munda languages, merges TAM categories with active and middle voices.

	Middle	Active
Present	=ta	=te
Present Progressive	=taˀjɖ	=teˀjɖ
Past Neutral (Past I)	=ki	=(y)oʔ
Irrealis	=na	=e
Past II	=khoʔ
Perfect	=siʔ(ɖ)
Optative	guɽuʔ/guɖuʔ

Causative

The causative derivation increases the valency of a contentive morpheme by introducing a higher or superordinate agent who causes the lower agent to act or a non-agentive event to happen. In Kharia, the signature marker of the Austroasiatic family -(o)(ʔ)b- (including allomorphs) is used as the causative prefix or infix. Double causative constructions are also allowed.

root	gloss	Simple causative	meaning	Double causative	meaning
aloŋ	'sing'	a-ˀb-loŋ	'have someone sing'	ob-a-ˀb-loŋ	'someone make someone sing'
ɖeˀb	'rise, climb'	o-ɖeˀb	'raise, offer up, sacrifice'	oˀb-ɖeˀb	'have someone sacrifice'
lemeˀɖ	'go to bed'	le-ʔ-meˀɖ	'put someone to bed'	oˀb-le-ʔ-meˀɖ	'have someone put someone to bed'
sore	'become ready'	so-ˀb-re	'prepare'	ob-so-ˀb-re	'have someone prepare'

Passive

The passive voice/reflexive in Kharia is realized as a standalone word ɖom, itself has no lexical meaning. Historically, it might have stemmed from the contentive morpheme dʒom ('to eat'), as it appears to cognate with Santali passive -jɔn (< jɔm 'to eat') and Sora-Juray reflexive/low transitive denoting marker -dəm-.

pothi=ki

book=PL

ter

give

ɖom=ki=may

PASS=MID.PST=3PL

pothi=ki ter ɖom=ki=may

book=PL give PASS=MID.PST=3PL

'(The) Books were given.'

Reciprocal

The Munda reciprocal *kƏl/*qəl is manifested in Kharia as a free-standing word kol, which has no lexical meaning. Whereas in other South Munda languages the reflexes of *kƏl exist as derivational prefixes and are used to derive new lexical verbs, the reciprocal marker in Kharia does not morphologically derive new words and does not directly change the meaning of the contentive morpheme of a TAM/Person-syntagma. In many cases, the interpretation of kol is broadly reciprocal, as illustrated:

ho=kiyar

that=DU

pothi=te

book=OBL

kol

RECP

lusu=ki=kiyar

grab=MID.PST=DU

ho=kiyar pothi=te kol lusu=ki=kiyar

that=DU book=OBL RECP grab=MID.PST=DU

'They grabbed each other book.'

In other contexts, it denotes what Lichtenberk (1985) described as collective, i.e. two participants did something together.

kol

RECP

aloŋ=ki=kiyar

sing=MID.PST=DU

kol aloŋ=ki=kiyar

RECP sing=MID.PST=DU

'They sang together.'

Lichtenberk also mentioned the third interpretation of the reciprocal is situation chaining, exemplified by the following sentence:

kol

RECP

buliˀj=taˀj=kiyar

follow=MID.PST=DU

kol buliˀj=taˀj=kiyar

RECP follow=MID.PST=DU

'They followed each other.'

Telicity

There are two telic markers in Kharia which serve the narrative structure:

• ɖoˀɖ (allomorph ɖoɽ) indicates that another event follows directly upon the event denoted by the predicate that it marks;

Tay

then

raja

king

jhaɽi

all

mudh

chief

kalo=ki=te

priest=PL=OBL

ro

and

yahudi

Jew

jait=aʔ

ethnic.group=GEN

etoŋɖag=aʔ

receiving.orders=GEN

guru=ki=te

teacher=PL=OBL

ɖokloʔ

meeting

remaʔ

call

ɖoʈh=oʔ

ATEL=ACT.PST

ro

and

ho=ki=te

that=PL=OBL

"masih=te

messiah=OBL

ate

where

jorme=na

be.born=INF

ayiˀj?"

QUAL.PRES

gam=oʔ

say=ACT.PST

ro

and

juŋ=oʔ

ask=ACT.PST

Tay raja jhaɽi mudh kalo=ki=te ro yahudi jait=aʔ etoŋɖag=aʔ guru=ki=te ɖokloʔ remaʔ ɖoʈh=oʔ ro ho=ki=te "masih=te ate jorme=na ayiˀj?" gam=oʔ ro juŋ=oʔ

then king all chief priest=PL=OBL and Jew ethnic.group=GEN receiving.orders=GEN teacher=PL=OBL meeting call ATEL=ACT.PST and that=PL=OBL messiah=OBL where be.born=INF QUAL.PRES say=ACT.PST and ask=ACT.PST

'The king called all the chief priests and teachers of the Jewish people who received orders [from the king] to a meeting and asked (said and asked) them "Where is the Messiah to be born?"'

• goˀɖ (allomorph goɽ) denotes a turning point or culmination in a narrative.

saitan=aʔ

Satan=GEN

fanda=te

net=OBL

bajhe

get.trapped

no

Q

goˀɖ=ta=iɲ

CTEL=MID.PRES=1SG

saitan=aʔ fanda=te bajhe no goˀɖ=ta=iɲ

Satan=GEN net=OBL get.trapped Q CTEL=MID.PRES=1SG

'Will I get trapped in Satan's net?'

Durativity: continuative, semel-iterative and iterative

In situations in which an action or event persists for a longer period of time than expected, or keeps on in action while performing another, the continuative marker kan is used.

ho=kaɽ

that=SG.HUM

kayom=ta

talk=MID.PRES

ro

and

kinbhar=jo

courtyard=ADD

joʔ

sweep

kan=te

CONT=ACT.PRES

ho=kaɽ kayom=ta ro kinbhar=jo joʔ kan=te

that=SG.HUM talk=MID.PRES and courtyard=ADD sweep CONT=ACT.PRES

'She talks and keeps sweeping the courtyard.'

With certain contentive morphemes, kan indicates that an action had been going on for a long time in the past before another action happened. With contentive morphemes denoting a volitional and motion event such as ɖoˀɖ 'take (away)', kan may express that the change is more-or-less permanent.

iɲ

I

am=bar=aʔ

2=2.HON=GEN

pothi=te

book=OBL

ɖoˀɖ

take

kan=oˀj

CONT=ACT.PST.1SG

iɲ am=bar=aʔ pothi=te ɖoˀɖ kan=oˀj

I 2=2.HON=GEN book=OBL take CONT=ACT.PST.1SG

'I took your book (and I'm not giving it back!).'

Semel-iterative, marked by loʔ, denotes an event that occurs one additional time or is repeated once. It is used primarily by elder speakers with masdars.

beroˀɖ

stand.up

loʔ=ki=may

SITER=MID.PST=3PL

ro

and

ɖoko

sit.down

loʔ=ki=may

SITER=MID.PST=3PL

beroˀɖ loʔ=ki=may ro ɖoko loʔ=ki=may

stand.up SITER=MID.PST=3PL and sit.down SITER=MID.PST=3PL

'They kept on standing up and sitting down again (e.g., at a meeting).'

With some contentive morphemes, loʔ does not seem to mark iterativity, but rather a prolonged event, making its meaning very similar if not the same to that of continuative kan:

gam~gam=loʔ=ki=may

RDPL~talk=CONT=MID.PST=3PL

gam~gam=loʔ=ki=may

RDPL~talk=CONT=MID.PST=3PL

'They kept in talking'

The main marker of iterativity in Kharia is khor. It indicates an action or event that comes to a stop and begins again, and may repeat in successive cycles.

ɖubhni=te

cup=OBL

peʔ

rice

yo=kon

see=SEQ

kowaʔ

crow

han=tiˀj

that=side

u=tiˀj

this=side

leŋ-leŋ

fly-RDPL

khor=ki

ITER=MID.PST

muda

but

soloʔ=te

dog=OBL

yo=kon

see=SEQ

peʔ

rice

ɲoʔ=na

eat=INF

um

NEG

are=ki

descend=MID.PST

ɖubhni=te peʔ yo=kon kowaʔ han=tiˀj u=tiˀj leŋ-leŋ khor=ki muda soloʔ=te yo=kon peʔ ɲoʔ=na um are=ki

cup=OBL rice see=SEQ crow that=side this=side fly-RDPL ITER=MID.PST but dog=OBL see=SEQ rice eat=INF NEG descend=MID.PST

'Having seen rice in the cup, the crow kept flying back and forth, but having seen the dog, it didn't come down to feed.'

Ambulative

The ambulative in Kharia is san (< sangoɖ 'walk', cf. Santali sɛn 'walk'), itself is no longer an independent lexical word. It denotes the action is carried out during movement from one place to another.

ol=na

bring=INF

olna

RDPL

kayar

mango

beloŋ

ripen

san=ki

AMB=MID.PST

ol=na olna kayar beloŋ san=ki

bring=INF RDPL mango ripen AMB=MID.PST

'While being brought [to market], the mangoes ripened along the way.'

When san is used as v2 for four contentive morphemes ol 'bring', ɖoˀɖ 'take', col 'go', and ɖel 'come', it rather denotes the sense of 'along with' instead of ambulativity. Kullu suggests that san might be compatible with the aforementioned four contentive morphemes, and in fact the first example is rejected by some speakers as ungrammatical.

baysaki

Baisali

tay

ABL

...ho=ki

that=PL

moɲ

one

baysaki

Baisali

koŋtaŋ

cow

ro

and

baysali

Baisali

sãɽ

ox

ol

bring

san=oʔ=ki

AMB=ACT.PST=3PL

baysaki tay ...ho=ki moɲ baysaki koŋtaŋ ro baysali sãɽ ol san=oʔ=ki

Baisali ABL that=PL one Baisali cow and Baisali ox bring AMB=ACT.PST=3PL

'From Baisali. they brought one Baisali cow and Baisali ox along with them.'

Totality

Totality, expressed by may, indicates completion or exhaustiveness of an action with respect to a particular participant of a scenario.

sori

together

sori

RDPL

chaha

tea

uˀd

drink

may=oʔ=ki

TOTAL=ACT.PST=PL

(*ro

and

tam

now

jou

up.to

uˀd=teˀj=ki)

drink=ACT.PROG=PL

sori sori chaha uˀd may=oʔ=ki (*ro tam jou uˀd=teˀj=ki)

together RDPL tea drink TOTAL=ACT.PST=PL and now up.to drink=ACT.PROG=PL

'They drank up all the tea together (*and are still drinking).'

Suddenity

Three markers bhaʔ, hambaʔ, and dhaˀb are used as v2 in TAM/Person-Syntagmas to denote an event or action that occurs suddenly or quickly. None of these words carry any lexical meanings.

ho=ki

that=PL

turthe=ga

fast=FOC

apan

REFL

apan

RDPL

oɽiyay

put.in.order

dhaˀb=na

SUD=INF

suru=yoʔ=ki

begin=ACT.PST=PL

ho=ki turthe=ga apan apan oɽiyay dhaˀb=na suru=yoʔ=ki

that=PL fast=FOC REFL RDPL put.in.order SUD=INF begin=ACT.PST=PL

'They quickly began to put themselves in order.'

Departive

The departive marker ʈu denotes a situation in which the subject performed some action and then left. This word, again, carries no lexical meaning.

maha

big

kulam

brother

paʔʈopur

Patna

tay

ABL

seŋ

first

muʔ=ki=may.

go.out=MID.PST=3PL

ho=ki

that=PL

kuɖaˀb

behind

aw=ki=may

stay=MID.PST=3PL

ho=ki=te

that=PL=OBL

u

this

gam

say

ʈu=yoʔ=ki

DPT=ACT.PST=PL

maha kulam paʔʈopur tay seŋ muʔ=ki=may. ho=ki kuɖaˀb aw=ki=may ho=ki=te u gam ʈu=yoʔ=ki

big brother Patna ABL first go.out=MID.PST=3PL that=PL behind stay=MID.PST=3PL that=PL=OBL this say DPT=ACT.PST=PL

'The elder brothers set out from Patna first. They said to those who were behind (before they left):"..."'

Conative

When an action is attempted by a participant without being succeeded or completed, the TAM/Person-Syntagma may employ dakha (less frequently lakha) to encode that type of modal-aktionsart. This word has no lexical meaning other than marking conativity to the predicate.

aɖ=ki=ko

ANAPH=PL=CNTR

mane

umh

ɖel=ki=may

come=MID.PST=3PL

arjowar

Aryavarta

gam

say

dakha=yoʔ=ki

CONAT=ACT.PST=PL

aɖ=ki=ko mane ɖel=ki=may arjowar gam dakha=yoʔ=ki

ANAPH=PL=CNTR umh come=MID.PST=3PL Aryavarta say CONAT=ACT.PST=PL

'They [i.e. others], uhm, came. They tried to call (say) [India] Aryavarta.'

Excessive

The Kharia equivalent to the English adverbial constructions involving 'strongly', 'intensely', 'extremely' is bay, itself homophonous and undoubtedly derived from the same contentive morpheme which means literally 'make, do'.

muˀjdaʔ

ant

feinga=te

grasshopper=OBL

leˀj

curse

bay=oʔ,

EXCES=ACT.PST

"u

this

chini

sugar

iɲ=aʔ

1SG=GEN

bohin=na=ɲ=ki

sister=1.POSS=1SG=PL

thuray=siʔ=may"

gather=PERF=3PL

muˀjdaʔ feinga=te leˀj bay=oʔ, "u chini iɲ=aʔ bohin=na=ɲ=ki thuray=siʔ=may"

ant grasshopper=OBL curse EXCES=ACT.PST this sugar 1SG=GEN sister=1.POSS=1SG=PL gather=PERF=3PL

'The ant gave the grasshopper a bad scolding, "This sugar my sisters have gathered."'

Autopoesis

In avolitional predicates, autopoetic marker jom denotes an action that happened on its own for the subject's own benefit, use, or maintenance, with the subject acting upon or for themselves. It may subtly imply an impolite connotation (i.e., act by oneself with no given permission). This word, although historically derived from an earlier contentive morpheme meaning 'eat', has become grammaticalized for autopoesis marking and is no longer found as an independent contentive morpheme.

iɲ

I

koʈhri=te

room=OBL

badli

change

jom=oˀj

AUTOPOES=ACT.PST.1SG

iɲ koʈhri=te badli jom=oˀj

I room=OBL change AUTOPOES=ACT.PST.1SG

'I just changed the room around (for no particular reason).'

Benefactive

Benefactive kay is used to denote that an action or event is performed for the benefit of another (external) participant. It does not carry any independent lexical meaning other than contributing the notion of 'do X on someone's behalf.' Peterson notes that kay is homophonous with the contentive morpheme kay which means 'to lift', but he also states that it appears unlikely to be the source of the benefactive marker.

khawna=ko

a.kind.of.sweet=CNTR

peʈhiya

market

geinka

every

ol

bring

kay=na

BEN=INF

laʔ=ki=kiyar

IPFV=MID.PST=3DU

khawna=ko peʈhiya geinka ol kay=na laʔ=ki=kiyar

a.kind.of.sweet=CNTR market every bring BEN=INF IPFV=MID.PST=3DU

'And [at] every market they would bring sweets for him.'

Partially-Finite and Non-Finite Predicates

Partially-Finite Predicates

In certain restricted types of daily discourse, such as events consisting of two sub-events that are highly related, there would be a non-final sub-predicate lacking some functional markers, often the person/number marker. Those partly-marked TAM/Person-Syntagma construction denoting co-subordination of events, that the following highly-connected action/event is being done by the same argument. They also mark the hierarchy of finiteness at clausal level, with the most finite TAM/Person-Syntagma receives full marking. The two sub-events may be directly juxtaposed next to each other or may be joined by a conjunction like ro or oɖoʔ.

onɖor=e

hear=ACT.IRR

ro

and

sikhe=na=niŋ

learn=MID.IRR=1Pl.INCL

onɖor=e ro sikhe=na=niŋ

hear=ACT.IRR and learn=MID.IRR=1Pl.INCL

'We should hear and learn [his teachings].'

Masdars

The unmarked bisyllabic and phonologically-motivated reduplicated form of a contentive morpheme or the masdar in Kharia can function in TAM/Person-Syntagmas as what is described as a "verbal noun" in Indo-European linguistics, i.e. a non-finite predicate that may also assume the referential and attributive functions.

no

CMPL

prabhu=waʔ

God=GEN

janam=te

birth=OBL

mane

umh

ho=ki=te

that=PL=OBL

iswar=aʔ

Lord=GEN

daŋ-daŋ

send-RDPL

aw=ki,...

QUAL=MID.PST

no prabhu=waʔ janam=te mane ho=ki=te iswar=aʔ daŋ-daŋ aw=ki,...

CMPL God=GEN birth=OBL umh that=PL=OBL Lord=GEN send-RDPL QUAL=MID.PST

'that at God's birth, umh, the Lord had sent to them [His Son] ([Jesus] was a sent-one of the Lord),...'

Infinitives

The infinitive marker in Kharia is =na, which appears to have emerged from the same origin with the middle irrealis =na. Peterson posits that the Kharia infinitive development may resemble what found in nearby Tibeto-Burman Newari in Nepal where a nonpast verbal form is also used to mark infinitive. The Kharia infinitive also seems to cognate with North Dravidian Kurukh infinitival form -na as well.

muda

but

dada=ɖom=ki=yaʔ

elder.brother=3.POSS=PL=GEN

aw-aw=te

QUAL-RDPL=OBL

ber

who

beʈi=ki=te

daughter=PL=OBL

um

NEG

uɖuŋ=na

make.flee=INF

pal=e

be.able=ACT.IRR

muda dada=ɖom=ki=yaʔ aw-aw=te ber beʈi=ki=te um uɖuŋ=na pal=e

but elder.brother=3.POSS=PL=GEN QUAL-RDPL=OBL who daughter=PL=OBL NEG make.flee=INF be.able=ACT.IRR

'But as long as their elder brothers are present, no one will be able to drive the daughters away.')

Sequential converbs

There are three sequential converb ("conjunctive participles") markers in Kharia: =ker, =ke, and =kon. The first two markers were borrowed directly from Sadri, and the last apparently derived from the contentive morpheme ikon ('make, do'), calqued from Indo-Aryan construction due to contact with Sadri and Hindi. These markers denote that sequential and temporal events being directly related and associated to one another in some way, combining to express a larger, complex event.

dhirom

slowly

dhirom

RDPL

khaɽiya

Kharia

maha

big

rokeˀd-loʔ

sand-land

paro=kon

cross=SEQ

rusuŋ

red

samuder

ocean

ɖam=ki=may

arrive=MID.PST=3PL

dhirom dhirom khaɽiya maha rokeˀd-loʔ paro=kon rusuŋ samuder ɖam=ki=may

slowly RDPL Kharia big sand-land cross=SEQ red ocean arrive=MID.PST=3PL

'Slowly but surely, the Kharia crossed the great desert and arrived at the Red Sea.'

There is a tendency that speakers would use sequential converbs to clauses which feature the same subject/agent (S/A). Although it is clearly a trend among speakers and sequential converbs can appear in different subject clauses as well, the latter use is considered uncommon. In fact, some consulted speakers reject the use of sequential converbs with different subject contexts as ungrammatical.

thor

few

toʔ

day

aw=kon

stay=SEQ

jiŋray=aʔ

porcupine=GEN

sori

with

ɖaŋ

send

goʈh=oʔ

CTEL=ACT.PST

thor toʔ aw=kon jiŋray=aʔ sori ɖaŋ goʈh=oʔ

few day stay=SEQ porcupine=GEN with send CTEL=ACT.PST

'[porcupine] having stayed a few days, [father] sent [daughter] off with the porcupine.'

etwa=te

Etwa=OBL

u=ki=yaʔ

this=PL=GEN

haleit

condition

yo=kon

see=SEQ

lebui

love

laʔ=ki

EMOT=MID.PST

etwa=te u=ki=yaʔ haleit yo=kon lebui laʔ=ki

Etwa=OBL this=PL=GEN condition see=SEQ love EMOT=MID.PST

'Etwa, seeing their condition, felt compassion.'

Imperfective converbs

Imperfective converbs (traditionally referred to as "present participles") in Kharia are used to indicate an event that happens simultaneously with main eventuality of the main finite predicate. There are four types of imperfective converbial constructions. The first three constructions employ marking on the semantic head of non-finite predicate and reduplication.

The genuine converb =ta, which might have derived from the present middle =ta, although the two are distinct in morphosyntactic behaviors.

ho=ki

that=PL

fawɖ

army

guˀɖ

like

katiˀb=ki=may

assemble=MID.PST=3PL

ro

and

hathiyar

weapon

ɖoˀɖ=ta

take=CVB

ɖoˀɖta

RDPL

seŋgor-ɖaʔ

straight-INT

raylogaɽh

Railogarh

hinte

LOC

caɽhai

attack

karay=oʔ=ki

do=ACT.PST=PL

ho=ki fawɖ guˀɖ katiˀb=ki=may ro hathiyar ɖoˀɖ=ta ɖoˀɖta seŋgor-ɖaʔ raylogaɽh hinte caɽhai karay=oʔ=ki

that=PL army like assemble=MID.PST=3PL and weapon take=CVB RDPL straight-INT Railogarh LOC attack do=ACT.PST=PL

'They assembled like an army, and, taking their weapons, [went] straight to Railogarh and there they attacked [the Kharia].'

Focus marker =ga can also functions as denoting semantic depictives:

ro

and

ho

that

kuda

millet

koloŋ=aʔ

bread=GEN

daru

tree

sumbhoʔ=te

base=OBL

iɲam=ga

cry=FOC

iɲamga

RDPL

goj

die

jom=ta

AUTOPOES=MID.PRES

ro ho kuda koloŋ=aʔ daru sumbhoʔ=te iɲam=ga iɲamga goj jom=ta

and that millet bread=GEN tree base=OBL cry=FOC RDPL die AUTOPOES=MID.PRES

'And crying and crying, she just dies [through sorrow] at the base of that millet bread tree.'

The infinitive marker =na may occur in imperfective converbial constructions.

...lekin

but

lam=na

search=INF

lamna

RDPL

souˀb=te

all=OBL

ikuɖ

very

jughay

much

daʔ

water

piyas

thirst

laʔ=ki

EMOT=MID.PST

...lekin lam=na lamna souˀb=te ikuɖ jughay daʔ piyas laʔ=ki

but search=INF RDPL all=OBL very much water thirst EMOT=MID.PST

'...But searching and searching, (they) all became very thirsty.'

Unmarked, repeated masdar and echo-formation of polysyllabic contentive morphemes can also play the role similar to that of imperfective converb markers since they denote the senses of durativity and iterativity, thus marking the secondary event as simultaneous within the time frame of the main event. Masdar and echo word imperfective converbs are restricted to polysyllabic morphemes.

modi

Modi

ro

and

saw-ɽay=ɖom

spouse-woman=3.POSS

ayo

mother

aba

father

ro

and

boker

brother.in.law

kulam=ɖom=ki=yaʔ

brother=3.POSS=PL=GEN

kaʈa

foot

sumboʔ=te

base=OBL

gur=kon

fall=SEQ

ĩyam

cry

ĩyam

RDPL

maf

forgiveness

bor=oʔ=kiyar

ask=ACT.PST=DU

modi ro saw-ɽay=ɖom ayo aba ro boker kulam=ɖom=ki=yaʔ kaʈa sumboʔ=te gur=kon ĩyam ĩyam maf bor=oʔ=kiyar

Modi and spouse-woman=3.POSS mother father and brother.in.law brother=3.POSS=PL=GEN foot base=OBL fall=SEQ cry RDPL forgiveness ask=ACT.PST=DU

'Modi and his wife fell down at the feet of his mother, father, brother-in-law and brothers, crying, asked for forgiveness..'

ho=ki

that=PL

poʔda=ɖom=ki=yaʔ

village=3.POSS=PL=GEN

halet

condition

onɖor=oʔ=ki,

hear=ACT.PST=PL

kunɖuʔ

child

hakon=ɖom=ki=te

ECHO=3.POSS=PL=OBL

sumtay=oʔ=ki

gather=ACT.PST=PL

ro

and

kuda

hurry

kudi

ECHO

khaɽiya

Kharia

ɖay=ki=yaʔ

woman=PL=GEN

kunɖaˀb

back

kunɖaˀb

RDPL

yar=oʔ=ki

flee=ACT.PST=PL

ho=ki poʔda=ɖom=ki=yaʔ halet onɖor=oʔ=ki, kunɖuʔ hakon=ɖom=ki=te sumtay=oʔ=ki ro kuda kudi khaɽiya ɖay=ki=yaʔ kunɖaˀb kunɖaˀb yar=oʔ=ki

that=PL village=3.POSS=PL=GEN condition hear=ACT.PST=PL child ECHO=3.POSS=PL=OBL gather=ACT.PST=PL and hurry ECHO Kharia woman=PL=GEN back RDPL flee=ACT.PST=PL

'They heard what was happening in the villages(the condition of their villages), gathered their children and hurriedly fled after the Kharia women.'

Participles

Kharia has two unambiguously participle markers, -l and =wala, both of which were borrowed from Sadri. The former only attaches to loan words of Sadri origin, e.g. biha 'marry', biha-l 'married'. It is restricted to past tense event only and has limited acceptance among speakers.

The second marker =wala is accompanied with infinitive =na. It is used to express iterative or habitual senses.

kamu=na=wala

work=INF=PTCP

lebu=ki

person=PL

kamu=na=wala lebu=ki

work=INF=PTCP person=PL

'working people'

The third, mysterious and native -ɖuʔ can be somewhat accepted as a participle marker, although it may be encountered occurring with contentive morphemes as a derivational marker carrying the approximative meaning of "-ish". E.g. maha-ɖuʔ 'biggish'. When this suffix functions as participle marker, there is no temporal orientation and the participle may refer to past, present or future events.

dilli

Delhi

co-ɖuʔ

go-PTCP

lebu

person

dilli co-ɖuʔ lebu

Delhi go-PTCP person

'the man, the person who went to Delhi'

Negation

The negative particles in Kharia cover the scope of TAM/Person syntagma, which is a syntactically clausal predicative structure. For primary predicates, there are two major types of negation, depending on the status of the TAM/Person-Syntagma. The table below summarizes types of negators found in Kharia.

Negator	Note
um	General; non-irrealis and non-imperative optative
abu	Imperative and optative
umboʔ	Emphatic

There is also a fourth negative marker na which has been borrowed from Indo-Aryan, however, has a very seldom use. It typically occurs in either negative commands, with the infinitive, or in a neither...nor expression with a finite TAM/Person-Syntagma:

Modi

Modi

na

neither

siloʔ=na

plow=INF

laʔ=ki

IPFV=MID.PST

no

or

oɖoʔ=ga

more=FOC

hoɖom

other

kamu

work

karay=na

do=INF

laʔ=ki

IPFV=MID.PST

Modi na siloʔ=na laʔ=ki no oɖoʔ=ga hoɖom kamu karay=na laʔ=ki

Modi neither plow=INF IPFV=MID.PST or more=FOC other work do=INF IPFV=MID.PST

'Modi neither plowed nor did any other work.'

General negative um

This negative marker always precedes the content head of now-partially finite TAM/Person-Syntagma. It pulls the subject marker away from its original attachment place, except the first (partially), dual and plural second person markers.

Simplified schematic overview of a negated TAM/Person-Syntagma:

NEG=PERS/NUM/HON RECP CAUS-Lexeme/CAUS/ v2=PERF=TAM/VOICE

umay

NEG.3PL

juŋ=oʔ

ask=ACT.PST

umay juŋ=oʔ

NEG.3PL ask=ACT.PST

'They didn't ask'

ubhroŋ

these.days

um=em

NEG=2SG

ɖe=na

come=MID.IRR

ubhroŋ um=em ɖe=na

these.days NEG=2SG come=MID.IRR

'These days you do not come.'

or

ubhroŋ

these.days

um

NEG

ɖe=na=m

come=MID.IRR=2SG

ubhroŋ um ɖe=na=m

these.days NEG come=MID.IRR=2SG

'These days you do not come.'

Um is also the negative contentive morpheme in general where it is placed before another contentive morpheme:

• Um lereʔ 'unhappy'; lereʔ 'joy, rejoice, joyous'
• Um suʔkhoʔ-boʔ 'uncomfortable'; sukhoʔ 'pleasure, happiness, happy' + -boʔ 'INT'

Modal negator abu

This negative marker is often used with prohibitive or optative constructions, although um may occur in the irrealis if the situation is not considered imperative. Note it has three version marked for person/number: abu '2SG', aʔbar '2DU/HON', aʔpe '2PL'. The form abu can also be used to refer to a 3rd person argument.

muda

but

soloʔ=te=ko

dog=OBL=CNTR

lemeɖ=ga

sleep=FOC

abu

NEG.MOD

ɖe=na

come=MID.IRR

muda soloʔ=te=ko lemeɖ=ga abu ɖe=na

but dog=OBL=CNTR sleep=FOC NEG.MOD come=MID.IRR

'But the dog just couldn't get any sleep' (but to the dog, sleep should not come).

Emphatic negative umboʔ

This negative marker is used in many situations, commonly in answers-to-questions to denote emphasis. It may occur at the end of clause after a finite TAM/Person-Syntagma, like the example below:

jiŋray

porcupine

diyo=ga

daily=FOC

laroʔ

date.palm

daru=te

tree=OBL

yo=na

see=INF

co=na

go=INF

laʔ=ki

IPFV=MID.PST

muda

but

goj=ta

die=MID.PRES

umboʔ

no

jiŋray diyo=ga laroʔ daru=te yo=na co=na laʔ=ki muda goj=ta umboʔ

porcupine daily=FOC date.palm tree=OBL see=INF go=INF IPFV=MID.PST but die=MID.PRES no

'The porcupine went to see the date-palm tree daily, but it just wouldn't die.

Umboʔ can occur with masdars as well, particularly in clauses that features high contextual density, where it may denote the sense of "could not." E.g. umboʔ col-col 'I could not go'. Consulted speakers state that in such contexts, only the first person arguments (i.e. the speakers) could be interpreted as the primary argument that the negated clause refers to. However, this interpretation appears to be only tendentially, as example from Pinnow (1965) shows that it does not carry a restricted reference to the first person arguments.

Qualitative Predication

Since Kharia lacks structural evidence for the existence of lexical categories of nouns and verbs, the term "copula" and its relative concept of nominal sentence are not applicable in this language. These concepts are subsumed under qualitative predications, which are divided into two types: stative qualitative predication ('to be') and dynamic qualitative predication ('to become'). The presence of qualitative predicative markers denotes a "narrative description", while the absence of those indicates "a general or self-evident truth".

In the stative qualitative predication category, there are three types of markers of stative qualitative predication: non-inherent, inherent, and neutral. The non-inherent markers express non-inherent, temporal, context-based, situational contingent qualities. The inherent markers express relatively inherent, defining properties to a referent. Both are limited to describe the current time only.

Type of stative qualitative predication	Non-negated form	Negated form
Inherent qualities	heke	nalage
Non-inherent qualities	ayiˀj(ɖ)	umboriˀj(ɖ)

Inherent qualities:

ho=kaɽ

that=SG.HUM

hosiyar

intelligent

heke

QUAL.PRES

ho=kaɽ hosiyar heke

that=SG.HUM intelligent QUAL.PRES

'She is the intelligent one.'

Non-inherent qualities:

ho=kaɽ

that=SG.HUM

hosiyar

intelligent

ayiˀj

QUAL.PRES

ho=kaɽ hosiyar ayiˀj

that=SG.HUM intelligent QUAL.PRES

'She is intelligent.'

Heke and nalage have been borrowed directly from Sadri. Malhotra (1982) cites some examples in which the negative form of heke is um heke instead of nalage. They are, however, quite rare, and nalage is considered standard.

Present-tense paradigm of non-negated, inherent qualitative marker

		singular	dual (HON)	plural
1st person	exclusive	hek=iɲ	heke=jar	heke=le
	inclusive		heke=naŋ	heke=niŋ
2nd person		hek=em	heke=bar	heke=pe
3rd person		heke	heke=kiyar	heke=ki/heke=may/hekãy

On the other hand, ayiˀj(ɖ) appears to be an Indo-Aryan loan, but which language it originated from is not clearly known. Peterson suggests a broadly Eastern Indo-Aryan origin, citing a cognate with the Maithili 3rd person non-honorific proform əich 'he/she/it'.

Present-tense paradigm of non-negated, non-inherent qualitative marker

		singular	dual (HON)	plural
1st person	exclusive	ayiˀjɖ=iɲ	ayiˀj=jar	ayiˀj=le
	inclusive		ayiˀj=naŋ	ayiˀj=niŋ
2nd person		ayiˀjɖ=em	ayiˀj=bar	ayiˀj=pe
3rd person		ayiˀj	ayiˀj=kiyar	ayiˀj=ki/ayiˀj=may

With the neutral stative qualitative predicative marker aw (derived from the same contentive morpheme meaning 'live, stay, remain'), the distinction between inherent and non-inherent qualities is disabled.

jhaɽi

all

poʔda

village

hinte

LOC

muɖu

one.HUM

la

then

muɖu

one.HUM

gãysi

spy

aw=ta=ki=ga

QUAL=MID.PRES=PL=FOC

jhaɽi poʔda hinte muɖu la muɖu gãysi aw=ta=ki=ga

all village LOC one.HUM then one.HUM spy QUAL=MID.PRES=PL=FOC

'In all villages there are [always] some spies.'

Kharia uses two contentive morphemes hoy and bone, both were borrowed from Sadri, to express the dynamic qualitative predication category ('to become', or more accurately, 'to take place').

u

this

jharkhanɖ

Jharkhand

alag

separate

hoy=ki

become=MID.PST

ho

that

bhere

time

tay

ABL

ghaɖ,

therefore

iskul

school

koʔghel=jo

vinicity=ADD

ho=ki

that=PL

khaɽiya

Kharia

kayam=ta=may

speak=MID.PRES=3PL

u jharkhanɖ alag hoy=ki ho bhere tay ghaɖ, iskul koʔghel=jo ho=ki khaɽiya kayam=ta=may

this Jharkhand separate become=MID.PST that time ABL therefore school vinicity=ADD that=PL Kharia speak=MID.PRES=3PL

'Since the time Jharkhand became independent, therefore they speak Kharia in school as well.'

ʈoʔpo

toppo.bird

aw=ki,

QUAL=MID.PST

adha

half

fura

part

niga

Kurukh

bone=ki=may

become=MID.PST=3PL

ʈoʔpo aw=ki, adha fura niga bone=ki=may

toppo.bird QUAL=MID.PST half part Kurukh become=MID.PST=3PL

'There was Toppo [a kind of bird], they [his desdendants, the Toppo clan] became half Kurukh.'

Sample text

Ughay

this.way

andai

guess

ɖom=ta

PASS=MID.PRES

no

CMPL

ata

Q

bhere

time

israeli

Israeli

lebu=ki

person=PL

rusuŋ

red

samudar=te

ocean=OBL

paro=na

cross=INF

laʔ=ki,

IPFV=MID.PST

hin

then

bhere=jo

time=ADD

khaɽiya=kia

Kharia=PL

ho

that

boʔ=ki=te=ga

place=PL=OBL=FOC

aw=ki=may.

live=MID.PST=3PL.SUBJ

Ughay andai ɖom=ta no ata bhere israeli lebu=ki rusuŋ samudar=te paro=na laʔ=ki, hin bhere=jo khaɽiya=kia ho boʔ=ki=te=ga aw=ki=may.

this.way guess PASS=MID.PRES CMPL Q time Israeli person=PL red ocean=OBL cross=INF IPFV=MID.PST then time=ADD Kharia=PL that place=PL=OBL=FOC live=MID.PST=3PL.SUBJ

'Thus it is assumed that at the time that the Israelis were crossing the Red Sea, at that time the Kharias were at those places as well.'

Laʔ

then

anin=aʔ

1PL.INCL=GEN

khariya

Kharia

bulbul,

Babylon

yane

i.e.

babilon,

Babylon

poʔda

village

tay

ABL

muʔ=kon

emerge=SEQ

del/em/-dol

come-?

arloʔ,

north

serloʔsin

south

ghay

way

hoy=kon

become=SEQ

utiˀj

this.side

del=ki=may.

come=MID.PST=3PL

Laʔ anin=aʔ khariya bulbul, yane babilon, poʔda tay muʔ=kon del/em/-dol arloʔ, serloʔsin ghay hoy=kon utiˀj del=ki=may.

then 1PL.INCL=GEN Kharia Babylon i.e. Babylon village ABL emerge=SEQ come-? north south way become=SEQ this.side come=MID.PST=3PL

'Then our Kharia [ancestors], having left Bulbul, i.e., Babylon, coming along, via the north and south, came to this side [i.e., here].'

Am=ga

You=FOC

patar

light

terter

give.RDPL

heke=m,

QUAL.PRES=2SG

am,

You

Yesu,

Jesus

ɖe=na=m

come=MID.IRR=2SG

ro

and

patar

light

ter=e=m

give=ACT.IRR=2SG

patar

light

ter=e=m.

give=ACT.IRR=2SG

Am=ga patar terter heke=m, am, Yesu, ɖe=na=m ro patar ter=e=m patar ter=e=m.

You=FOC light give.RDPL QUAL.PRES=2SG You Jesus come=MID.IRR=2SG and light give=ACT.IRR=2SG light give=ACT.IRR=2SG

'You are the [one] who gives light, you, Jesus, will come and give light, you will give light.'

References

Bibliography

• Hengeveld, Kees; Lier, Eva van (2010). "Parts of speech and dependent clauses in Functional Discourse Grammar". In Ansaldo, Umberto; Dan, Jon; Pfau, Roland (eds.). Parts of Speech: Empirical and theoretical advances. John Benjamins Publishing Company. pp. 253–285. doi:10.1075/bct.25.12hen. ISBN 978-9-02722-255-8.
• Lier, Eva van (2009). Parts of Speech and Dependent Clauses: A Typological Study. LOT. ISBN 978-9-07832-8-957.
• Lier, Eva van, ed. (2023). The Oxford Handbook of Word Classes. Oxford University Press. doi:10.1093/oxfordhb/9780198852889.001.0001. ISBN 978-0-19188-7-185.
• Peterson, John M. (2007). "Languages without nouns and verbs? An alternative to lexical classes in Kharia". In Masica, Colin P. (ed.). Old and new perspectives on South Asian languages: grammar and semantics. Papers growing out of the Fifth International Conference on South Asian Linguistics (ICOSAL-5), held at Moscow, Russia in July 2003. Delhi: Oxford University Press. pp. 274–303. ISBN 978-8-12083-208-4.
• Peterson, John (2008). "Kharia". In Anderson, Gregory D. S. (ed.). The Munda languages. Routledge Language Family Series. Vol. 3. New York: Routledge. pp. 434–507. ISBN 0-415-32890-X.
• Peterson, John M.; Maas, Utz (2009). "Reduplication in Kharia: the masdar as a phonologically motivated category". Morphology. 19: 207–237. doi:10.1007/s11525-009-9141-x.
• Peterson, John M. (2009). A Kharia-English Lexicon [HL Archive 5]. UC Santa Barbara. doi:10.5070/H90023672.
• Peterson, John M. (2010). Kharia Texts, translated, glossed, and annotated by John Peterson. Osnabrück University.
• Peterson, John M. (2010). "Language contact in Jharkhand: Linguistic convergence between Munda and Indo-Aryan in eastern-central India". Himalayan Linguistics. 9 (2): 56–86. doi:10.5070/H99223479.
• Peterson, John M. (2011). A Grammar of Kharia: A South Munda Language. Brill. ISBN 978-9-00418-720-7. OCLC 31045692.
• Peterson, John M. (2013). "Parts of speech in Kharia: a formal account". In Rijkhoff, Jan; Lier, Eva Helena van (eds.). Flexible word classes: typological studies of underspecified parts of speech (1 ed.). Oxford: Oxford University Press. pp. 131–168. doi:10.1093/acprof:oso/9780199668441.003.0005. ISBN 978-0-19-966844-1.
• Peterson, John M. (2014). "Figuratively speaking - number in Kharia". In Storch, Anne; Dimmendaal, Gerrit J. (eds.). Number – Constructions and Semantics: Case studies from Africa, Amazonia, India and Oceania. John Benjamins Publishing Company. pp. 77–110. doi:10.1075/slcs.151.04pet. ISBN 978-9-02725-916-5.
• Peterson, John M. (2017). "Jharkhand as a 'Linguistic Area': Language Contact Between Indo-Aryan and Munda in Eastern-Central South Asia". In Hickey, Raymond (ed.). The Cambridge Handbook of Areal Linguistics. Cambridge University Press. pp. 551–574. doi:10.1017/9781107279872.021.
• Peterson, John M. (2023). "Lexical and Grammatical Categories in RRG". In Bentley, Delia; Usón, Ricardo Mairal; Valin, Robert D. Van; Nakamura, Wataru (eds.). The Cambridge Handbook of Role and Reference Grammar. Cambridge: Cambridge University Press. pp. 181–217. doi:10.1515/9783110260021.89. ISBN 978-1-00935-355-7.

Further reading

• Gagan Chandra Banerjee (1894). Introduction to the Khariā Language. Bengal Secretariat Press. ISBN 9788120617728. Retrieved 25 August 2012. {{cite book}}: ISBN / Date incompatibility (help)

External links

• Online Kharia Dictionary Archived 2016-03-03 at the Wayback Machine