Krishnamurti Bhadriraju. The Dravidian Languages

Подождите немного. Документ загружается.

6.3 Cases 241

AKu. -iddu is from irdu ‘having been’. Kota markers -tr/-ltr are apparently not derived

from the above, e.g. ayk-tr ‘from there’, pay-ltr ‘from the house’. Purposive -k¯a

ry of

obscure origin is a postposition, e.g. Ko. n¯aym-t-k¯a

ry ‘for justice’. Toda ablative -ni

d is

derived from

∗

-nintu; it also has -sn whose origin is not known, e.g. kw¨ıg-sn ‘from the

plains’. Locative -kiz/-giz are postpositions.

dagu: the postpositions o

d¨ı (cf. Ta.–Ma. ko

tu) ‘having taken’ and kay-˜nja

∗

kay-ind-a) ‘by hand’ are used in the instrumental meaning, e.g. kat-tin-o

d¨ı ‘with

a knife’, n¯a

da-kay ˜nja ‘by me’ (cf. Te, c

¯eta ‘by hand’). In the sociative meaning jot¨eli,

k¯u

da and pakka are used as postpositions, e.g. ave

da k¯u

da ‘with him’; in the ablative

the postpositions are i ˜nji (<

∗

irnj- < irundu) and kay

nja, e.g. nann i ˜nji ‘from me’,

mara-t-i ˜nji ‘from the tree’, n¯a

da kay ˜nja ‘from me’ (lit.‘from my hand’). Locative -alli

which is apparently borrowed from Kanna

da is a grammaticalized postposition, mean-

ing ‘in’.

Kanna

da: clearly the locative u

l/o

l, o

lage and alli are postpositions, puyyal-o

l ‘in

distress’, enn-alli ‘with me’. The locative postposition is also used in instrumental and

causative meanings.

lu: the ablativ

e case markers, -

ntu and -

dd¨ı look like truncated

words of unknown

identity, ka

do-ntu ‘from the ﬁeld’, mara-

dd¨ı ‘from the tree’.

South Dravidian II Telugu: two non-ﬁnite verbs, k¯urc-i (k¯ur-cu ‘to put together’ tr

of k¯u

du ‘to meet’), gu

rinc-i (gurincu ‘to aim at’, a frozen verb) ‘having regard to,

concerning, about’, are added to inﬂected accusative nouns, ¯ar¯aju-nu g¯urci/gu

rinc-i

tini ‘I heard about that king’. These occur both in Old and Modern Standard Telugu.

The causee agent is marked by the instrumental postposition c¯et-an, c¯e-(a)n ‘by’ (the

instrumental of c¯e(yi) ‘hand’), the sociative–instrumental is marked by t¯o-(n), t¯o

d-an

(t¯o

du ‘company’), katti-t¯o ‘with a knife’, s¯ıta-t¯o ‘with Sita’, Three postpositions are

used with dative meanings, namely OTe. kai ‘for’ (← ku + ayi ‘having been for’),

OTe. ko

ra-ku ‘for the purpose of ’ (kora ‘purpose, work’), OTe. po

te (< SD

∗

por-

t-) ‘for the purpose of’, cf. Ta. poru

tu ‘for the sake of ’, poru

l ‘thing, matter’

[DEDR 4544],

Mdn Te.

k¯osam/k¯osar

‘for

the sake of

’ (perhaps

from dat.

ku +

¯osara < Skt. avasara- ‘pupose’). In the ablative u

d-i (ppl of u

du ‘to be’) is added

to the accusative in -n; consequently, by reanalysis, nu

di became the grammaticalized

postposition meaning ‘from’. Modern Telugu uses nunci/ninci (ppl uncu ‘to keep’, tr of

du ‘to be’) but interpreted as an intransitive, e.g. ¯ame ¯uri-ninci waccindi ‘she came

from the village’ (literally, ‘being in the village, she came’). Old Telugu also used k¯olen

‘for’. Another widely used postposition in the sense of ‘because’ was walana(n) >

walna(n) > walla from walanu ‘possibility, skill’, e.g. n¯a walla ‘because of me’; others

include d¯aka(n) ‘up to, until’ > Mdn Te. d¯ak¯a : i

tid¯ak¯a ‘up to the house’, ninna

ti-d¯ak¯a

‘until yesterday’.

242 Nominals

The genitive was marked by a postposition of obscure origin, -yokka, in learned speech

and in commentaries on Sanskrit texts, but not in poetic compositions. This could be

related to PD

∗

o ‘to join, unite’ which had an inﬁnitive -okka in Tamil and Malay¯a

lam.

This is also a case of a frozen, inﬂected verb. It is not also used in Modern spoken

Telugu and is considered unidiomatic in writing. In the locative, Old Telugu has andu

‘that place’, -l¯o(n), -l¯opala(n) ‘inside’, of which the last two continue in Modern Telugu.

Several words of direction are also used, namely mundu ‘front’, wenaka ‘back’, daggara

‘near’, m¯ıd-a, pai-na ‘above’, kind-a, diguwa-na, a

dugu-na ‘below, at the

bottom

’. The

comparative postposition

is -

te (said to be dati

ku + a

t¯e ‘to’ + ‘if one says’) both

in Old and in Modern Telugu, e.g. w¯a

du n¯a-ka

te po

dugu ‘he is taller than me’ (lit. ‘he-I

obl more than- tall’).

Sometimes more than one postposition can be used, e.g. i

t(i) l¯o-nunci ‘from in-

side the house’, ce

tu daggar-i-ki ‘near (to) the tree’ (also see Mahadeva Sastri 1969:

166–70).

Gondi: -agg¯a ‘with’ occurs after human nouns, v¯on agg¯a ‘with him’. The Koya dialect

uses t¯o

te ‘with’ for instrumental-sociative, e.g. go

del-t¯o

te ‘with an axe’, n¯a-t¯o

‘with me’. Tyler (1969: 54–5) calls this a sufﬁx and not a postposition, although it looks

like a clear borrowing from Early Telugu oblique of t¯o

du (perhaps originally

∗

t¯o

du),

namely t¯o

ti. He gives a list of ﬁfteen postpositions, e.g. porro ‘on’, perke ‘after’, pakka

‘beside’, kunci/nunci ‘from’, etc.

da: the postpositions include vale ‘in company with’, vandi

‘for the purpose of ’,

daka ‘until’ (< lw Telugu), lo

i ‘inside’, ban ‘place’, musku ‘above’, a

dgi ‘below’, etc.

(Krishnamurti 1969a: 261).

Kui: -

rahi marks the instrumental, heni-

rahi ‘with a razor’. Winﬁeld lists (1928: 24–

5) a number of postpositions for dative, ablative and locative, classiﬁed as (a) location

(locative case meanings), (b) motion (to and from), and (c) association (with, by means

of) etc. Examples: tani ‘in’, lai ‘below’, s¯o

ri ‘near’;

rai ‘from’, bahata-ngi ‘motion to’

(humans); tangi ‘for’, tingi ‘because of’, rohe ‘together with’,

rai ‘with’ (instrumental).

Kuvi: the postpositions include ta

na ‘at’ (loc), ta

na -

ti ‘from place’ (abl), tole/tale ‘in

company with’ (soc), and seventeen forms in the locative, e.g.

dagre ‘near’, dari ‘near’,

t¯ale

ni ‘below’, two for ‘limitative’, epe ‘up to, until’, and three for comparative, e.g.

kih¯a ‘than’, leh¯e ‘like’ (Israel 1979: 73–6).

Pengo: the list includes laha

, huda

, hoke in instrumental meaning, sa

as a socia-

tive marker. The ablative is marked by n¯a

‘since’, also used as instrumental with

some nouns, homb¯an¯a

‘since Monday’, k¯ısa

n¯a

‘with tusks’. For dative, k¯aji

and bisre ‘for the sake of” are used. Six postpositions are given for the locative, e.g.

t¯ake, bitre, m¯ıgo, h¯agi, etc. ¯ezu

t¯ake ‘into the water’, mar jopi ‘on to the tree’, mar h¯agi

‘underneath a tree’ (Burrow and Bhattacharya 1970: 41–3).

6.4 Pr

onouns

243

Central Dravidian Kolami: in the ablative meaning -tana

t,-natta

d,-na

d ‘from,

through’ are used, e.g. ¯ur tana

t ‘from the village’. In the dative -nada

/-ada

‘to, near’

are used, e.g. amn-ada

‘his possession’.

Naik

ri uses -l¯a

d and -tan¯a

t for the ablative, ell¯a tan¯a

t ‘from the house’, phoy l¯a

‘from above’.

Naiki (Chanda): for the sociative meaning, -nokon is used: t¯ole-nokon ‘with father’,

maye-nokon ‘with mother’. In the locative -lopun occurs, e.g. ¯ı r -lopun ‘inside water’.

Parji: in the locative the postposition -kan/-ka (cf. OTa. ka

n) occurs after the genitive

of a noun, followed by the ablative sufﬁx -

ti/-

ta, e.g. murtal-in-kan

ta ‘from (the place

of) the old woman’, mer-to-ka ‘on to the tree’.

Ollari: -payi

t/-pay

ti is added to the inﬂected dative, e.g. ayal-i

payi

t ‘for (the sake

of) mother’, k¯ondel-i

payi

t ‘for the cows’. -pel

is used in the ablative meaning, e.g.

mar-in-pel

‘from the tree’ (Bhattacharya 1957: 22).

Gadaba: the list includes a

ter ‘away from’, a

dgun ‘below’, kakel ‘near’, ta

drel ‘in-

side’, kana ‘than’,

d¯a

ka ‘till’ etc. (Bhaskararao 1980: 60–1).

North Dravidian Ku

rux: the instrumental sufﬁxes look like postpositions, i.e. -tr¯ı/-tr¯u

and the adverb l¯ek¯e. So does the genitive gahi ‘possession’.

Malto: the sociative is marked by -gusan/-gan¯e, ¯alar-gu

thiar-gusan ‘with men’.

Brahui: it is not certain how to distinguish case sufﬁxes from postpositions morpho-

logically.

Of the locati

ve, postposition

‘inside’ [

∗

l-a- 698] can be reconstructed for Proto-

South Dravidian because it occurs in Kanna

da in SD I and Telugu, Gondi, Kui–Kuvi of

South Dravidian II. Many of the native words used as postpositions can be reconstructed

for Proto-Dravidian like the words meaning ‘above, high’ [

∗

m¯e,

∗

m¯e-l 5086] and ‘below’

[

∗

k¯ı

z 1619]. I have not attempted it for all the postpositions.

6.4 Pronouns

Pronouns anaphorically refer to nouns and can be substituted

for them. They are a sub-

class

of nominals since they are distinguished for number and/or gender

and carry case

markers. Pronouns are divided into three subclasses: (1) personal pronouns, (2) demon-

strative pronouns and (3) interrogative pronouns. Personal pronouns occur in ﬁrst and

second persons and in the reﬂexive. These are distinguished only for number and not

for gender. Demonstrative and interrogative pronouns are distinguished for number as

well as gender. Etymologically also there are differences between these subclasses. The

demonstrative pronouns are derived from deictic roots meaning ‘this, that, yonder’, while

the interrogative pronouns are derived from the interrogative root, meaning ‘what?’ or

‘who?’ The personal pronouns are primary forms and are not derived from other roots.

244 Nominals

able 6.5a.

irst-person-singular

pronouns

in the

nominative and oblique

Group I: First-person-singular ‘I’ [DEDR 5160]

Nominative Oblique

Proto-Dravidian

∗

yaH-n/

∗

y¯a-n

∗

yan-

Ta. y¯a

nen-

Ma. — en-

To. ¯on en-

Ko. ¯an en-, e-

d. — en-

Ka. ¯aN

en-

Tu. (dial) y¯an¨ı, ¯en¨ı (dial) en-

Te. ¯enu —

Go. (dial) an¯a, ann¯a, ana —

Kui ¯anu —

Pe. ¯an(e

) —

Man

da ¯an —

Kol. ¯an an-

Nk. ¯an an-

Pa. ¯an an-

Oll. ¯an an-

Gad. ¯an an-

r. ¯en e

Malto ¯en e

Br. ¯ı (?) kan-

Morphologically all pronouns carry case markers and postpositions. Interrogative and

demonstrative pronouns are used syntactically in correlative constructions of the type,

‘whoever asked ...he’ (section 9.3.2.3).

6.4.1 Personal pronouns

The personal pronouns pose interesting phonological and morphological problems in

Dravidian. We will start with the cognates of the pronouns in the ﬁrst person in different

Dravidian languages.

6.4.1.1 The ﬁrst person

Tables 6.5a–e give the forms of the ﬁrst person in the nominative and oblique divided

into phonologically coherent groups.

In Kanna

da, morphophonemic//N, M// are phonemically /m/ ﬁnally, but /n/ and /m/ respectively

in gemination. Intervocalically N = /n/, M = /m/ or /v/. Apparently PD

∗

n fell together with

∗

in Early Kanna

da in the utterance-ﬁnal position, but the contrast is preserved elsewhere, see

Ramachandra Rao (1972:

§§2.7, p. 38).

6.4 Pr

onouns

245

able 6.5b.

Alternative

forms

of the

ﬁr

st person singular

Group II Nominative Oblique

Proto-South Dravidian

∗

˜n¯an-

∗

nan-,

∗

˜n¯a-

Ta. n¯an (> Mdn n

a )—

Ma. ˜n¯an (dial n¯an)

—

Kod. n¯an¨ı, n¯a nan-, n¯a-

Ka. n¯aN nan-

Te. n¯enu; n¯anu (mdn

regional substandard) nan-, n¯a-

Go. (dial) na(n)n

¯a, nan n¯a-

Konda n¯an n¯a-

Kui (BB) n¯anu nan-, n¯a-

Kuvi n¯an¯u (F), n¯anu (S) n¯a-

Pe. — na

g, n¯a-

It is clear that Groups I and II are not etymologically related. Group I is derived

from Proto-Dravidian

∗

y¯an/

∗

yan-

and Group II, which is conﬁned only to South

Dravidian I and South Dravidian II, requires us to reconstruct

∗

˜n¯an/

∗

˜nan- at the Proto-

South Dravidian stage. We will see below what led to the creation of a second form in the

singular in Proto-South Dravidian. I discussed the problems of the personal pronouns

in great detail in an article in 1968 (reprinted with a postscript in Krishnamurti 2001a:

76–98), which continue to be discussed by Dravidianists. It is well established that the

vowels

∗

¯a/

∗

¯e alternate after the reconstructed palatal consonants

∗

y and

∗

˜n. The presence

∗

˜n- in Malay¯a

lam and the alternation of the vowels in Telugu are sufﬁcient to posit

∗

˜n- in Proto-South Dravidian for Group II. Also notice that the oblique is formed by

shortening the long vowel of the nominative in both Group I and Group II. Additionally,

in Group II there

is a second oblique normally used in non-accusative cases.

It is formed

by dropping the ﬁnal consonant of the nominative.

There are no serious phonological problems. There was neutralization of contrast

between

∗

¯a and

∗

¯e in Proto-Dravidian after

∗

y leading to an alternation between these

two vowels in different languages. South Dravidian I has ¯a-/e- (nominative/oblique),

A. C. Sekhar (1953: 88–9) quotes one instance of Ma. n¯an in a tenth-century inscription and

suggests, in agreement with Ramaswami Aiyar, that ‘ ˜n¯an may have been a west coast archaism’.

In old and modern Malay¯a

lam ˜n¯an is the standard form of the 1sg; only ˜n¯an (but not n¯an) occurs

in R¯amacaritam, the earliest literary work in Malay¯a

lam (c. twelfth century AD); see George

(1956: 159).

My recent research supports the positing of a laryngeal

∗

H in the personal pronoun in Early

Proto-Dravidian as

∗

yaHn/

∗

yan-, where aH becomes ¯a in Proto-Dravidian when the outcome is a

free form; the laryngeal is lost, when the outcome is a bound form like

∗

yan-. Since this proposal

is still to be discussed by Dravidian scholars I am giving only the traditional reconstructions in

this chapter.

246 Nominals

able 6.5c.

irst-person-plur

al pronouns

in the nominative

and oblique

Group III: First-person-plural ‘we’ (exclusive, unless otherwise speciﬁed)

[DEDR 5154]

Nominative Oblique

Proto-Dravidian

∗

yaH-m/

∗

y¯a-m

∗

yam-

Ta. y¯am; y¯a˙n-ka

em-

Ma. — e˙n-˙nal-

To. em em-

om (incl) om- (incl)

Ko. ¯am em-

am- (incl)

Kod. e

-ga e

-ga-

Ka. ¯aM em-

Tu. (dial) y¯a˙n-ku

l¨ı, e˙n-ku

l¨ıya˙n-ku

le-, e˙n-ku

le-

Te. ¯emu —

Go. amm-

¯a

t, amm-o

t, amm-ok —

Kui ¯amu —

Pe. ¯ap(e

) —

Man

da ¯am —

Kol. ¯am am-

Nk.(Ch.) ¯am am-

Pa. ¯am am-

Oll. ¯am am-

Gad. ¯am am-

r. ¯em em-

Malto ¯em em-

Central Dravidian ¯a-/a- and North Dravidian ¯e-/e-. Tu

lu and Telugu have ¯e-, while the

other South Dravidian II languages have ¯a- in the nominative. Old Tamil preserved the

∗

y- but it was replaced by Ø- in Later Tamil and all other languages.

Toda ¯o-

corresponds to Ta. ¯a-. What is important to note is that the South Dravidian II languages

have no phonologically related oblique forms in Group I. The oblique bases of these

languages have initial n- (see Group II).

Table 6.5c gives the plural forms. There were in Proto-Dravidian two plurals in the

ﬁrst person, one including the person addressed, called the inclusi

ve plural, and the other

excluding the person addressed, called the exclusive plural. Table 6.5c mainly refers to

forms of the ﬁrst person exclusive, but there was already confusion between the exclusive

The Tu

lu non-Brahmin dialect has y¯an¨ı, which appears to be a recent development by lowering

thevowelin(y)¯en¨ı. There are no other cases of retention of PD

∗

y-inTu

lu to support the

assumption that this was a retention.

‘The double plurals n¯a˙nka

l and y¯a˙nka

l are found as early as the C¯ıvakacint¯ama

ni period’

(Subbaiya 1923: 1).

6.4 Pr

onouns

247

able 6.5d.

Alternative

forms

of the

ﬁr

st person plur

al (exclusive)

Group IV Nominative Oblique

Ta. n¯a˙n-ka

l (Mdn n¯a˙n-ka/-ka

l)—

Ma. ˜n¯a˙n-˙na

l ˜na˙n-˙na

Kod. na

-ga na

-ga-

Te. n¯emu, m¯emu mamm-, m¯a-

Go. (dial) mamm-

¯a

t, m¯a-

m¯am-a

t, mamm-o

mamo-o, m¯a-

mar-at, mamm-a, m¯a-m

da m¯ap m¯a-

Kui (BB) m¯amu ma-

Kuvi m¯amb¯u (F), m¯ambu (S)

m¯arr¯o (F), m¯aro (S) m¯a-

Pe. — ma

g-, m¯a-

Table 6.5e. The ﬁrst-person-plural (inclusive) forms

Group V: First-person-plural ‘we’ (inclusive, unless

otherwise speciﬁed) [DEDR 3647]

Nominative Oblique

Ta. n¯am nam(m)-

Ma. n¯am nam(m)

Ka. n¯aM(n¯am, n¯avu) nam(m)-

Tu. nama nama-

Te. manamu mana- (? < nam-a-)

[Go. namo

t (excl) m¯a-]

Kol. n¯en

dn¯en

Nk. n¯en

d/n¯em n¯en

r. n¯am n¯am-, na

-g-

Malto n¯am nam-

[Br. nan (incl and excl) nan-]

and inclusive plurals. The forms in Group III which

are derivable from

∗

y¯am/

∗

yam-

denote the ﬁrst person exclusive. Note that the oblique is formed uniformly by replacing

the long vowel of the nominative by a corresponding short vowel (with expected variation

following

∗

y).

The forms in Group IV and Group V are obviously related. Group V forms represent

the inclusive plural but no language shows

∗

˜n-; however, the alternation of ¯a/¯e (see

Te. n¯emu, m¯emu and n¯e- forms in Kolami and Naiki) suggest an original

∗

˜n-. Actually

Malay¯a

lam preserves the ˜n- forms in Group

IV although the meaning is

‘we (exclusive

)

’.

It is clear that the distinction between the inclusive and exclusive forms got disturbed

248 Nominals

mainly in the case of forms with a nasal initial. The question is how does Group IV

mean ‘we (exclusive)’ and Group V ‘we (inclusive)’? In the literary languages also the

distinction between the ‘inclusive’ and ‘exclusive’ got blurred from early times. I made

the following proposal to account for all irregularities including the m-initial forms in

the languages of South Dravidian II.

In Proto-Dravidian there was one singular form

∗

y¯an/

∗

yan- ‘I’ and two plural forms,

∗

y¯am/

∗

yam- ‘we (exclusive)’ and

∗

˜n¯am/

∗

˜nam- ‘we (inclusive)’. These are preserved

intact in Central Dravidian (Kolami–Parji etc.) and North Dravidian (Ku

rux–Malto).

In Proto-South Dravidian, the parent of South Dravidian I and II, a second singular

∗

˜n¯an/˜nan- was analogically created through back formation from the second plural

∗

˜n¯am/˜nam-. Only

these languages show

n- forms

beside the initial zero forms inherited

from Proto-Dravidian

∗

y¯an/

∗

yan-. The formal rivalry between the two sets of forms,

not supported by a functional (semantic/grammatical) difference, has further led to a

number of changes:

(a) Note formal mismatch between the nominative and oblique forms (see the gaps

represented by dashes for Groups I to IV). Such a mismatch is conﬁned to those languages

that have inherited the innovative singular in n- beside the normal form in Ø-.

(b) The presence of two singulars (with no meaning difference) and two plurals

(with meaning difference) has weakened the original meaning contrast between the

two plural

forms. The addition of -

l to the inclusi

ve/exclusive plural

∗

˜n¯am/

∗

y¯am (see

Groups III and IV) to mean an exclusive was a shared feature of Tamil–Malay¯a

lam–

Iru

la– Ku

rumba–Ko

dagu of South Dra

vidian I. Kanna

da had lost the inclusi

–exclusive

distinction when one set got normalized as n¯an: n¯am replacing

∗

¯an/

∗

¯am. In South

Dravidian II, the inherited plural

∗

n¯am/

∗

n¯em, which later developed to m¯am-/m¯em, be-

ing restructured on the analogy of the oblique

∗

m¯a-, meant only ‘we (exclusive)’ (see

by the ﬁnal consonants, -n for singular and -m for plural. But when the languages of

Proto-South Dravidian II innovated a second oblique (in cases other than nominative

and accusative) with the loss of the ﬁnal consonant, i.e.

∗

˜n¯an 

∗

˜n¯a- >

∗

n¯a-,

∗

n¯am 

∗

n¯a->

∗

n¯a-, the singular–plural contrast was lost. It was restored by a morphological

change of substituting m- for n- initially in the oblique plural of all these languages

uniformly. Note that all languages of South Dravidian II have n¯a- as the oblique stem

for the nominative n¯an/n¯en in the singular (‘I’), and m¯a- as the oblique stem in the

plural (‘we’) corresponding to the nominative

∗

n¯am/

∗

n¯em. Later, the nominatives got

restructured on the basis of the oblique stems; consequently we notice variation between

n- and m- stems in the nominative of the plural (study Groups IV and V).

(d) Different languages of South Dravidian II have independentlyrestored the inclusive

plural, not related to the inherited set, by morphological innovation, e.g. Te. mana-mu

6.4 Pr

onouns

249

able 6.6a

Second-per

son-singular

pronouns

in the

nominative and oblique

Group I: Second-person-singular ‘thou’ (DEDR 3684)

Nominative Oblique

Ta. n¯ı ni n(n)-

Ma. n¯ı nin(n)-

To. n¯ı n ¨ın-

Ko. n¯ı nin-/ni-/di-

Kod. n¯ı n ¨ı, n¯ı nin-, n¯ı-

Ka. n¯ıN, n¯ın(u) nin(n)-

Tu. ¯ı nin-, in-

Te. n¯ıwu, ¯ıwu nin-, n¯ı-

Go. — n¯ı -

da n¯ınu ni

-, n¯ı-

Kui ¯ı n u (W) n¯ı-

n¯ınu (BB) nin-, n¯ı-

Kuvi n¯ı n ¯u (F), n¯ınu (S) n¯ı -

Pe. ¯en(e

) ni

- (acc–dat), n¯ı -

Kol. n¯ıv in-

Nk.(Ch.) n¯ı v, ¯ıv in-

Pa. ¯ın in-

Oll. ¯ın in-

Gad. ¯ın in(n)-

r. n¯ın ni

Malto n¯ı n ni

Br. n¯ı ne-, n-

(?<

∗

nama-), Go. apl¯o (lw < Marathi), Ko

da m¯a-

t (contrasting with exclusive m¯a-p;

t occurs as plural sufﬁx in Gondi but of obscure origin), Kui ¯aju,Pe.¯as/¯ah (< ?¯a

t-V-).

A detailed discussion of the arguments and counter arguments relating to the above

issues with a critical review occurs in Krishnamurti 1968: 189–205, 2001a: 76–98 (see

particularly the postscript).

6.4.1.2 The second person

Group I in table 6.6a deri

ves from the reconstruction

∗

n¯ın/

∗

nin-. The loss of ﬁnal -n

in the nominative is a shared innovation of a subgroup of South Dravidian I (Tamil,

Malay¯a

lam, Ko

dagu, Toda, Kota and Tu

lu); that there was a ﬁnal -n is attested by its

presence in the oblique form. The initial n- is lost in all Central Dravidian both in the

nominative and in the oblique as a shared innovation. The loss is also found sporadically

in Tu

lu, Telugu and Pengo. The ﬁnal -n is replaced by -wu in Early Telugu, but the

oblique is nin-. Kolami–Naiki have apparently borrowed the nominative forms from

250 Nominals

able 6.6b

Second-per

son-plur

al forms

Group IIA: Second-person-plural ‘you’ (DEDR 3688)

Nominative Oblique

Ta. n¯ım, n¯ı˙n-ka

l, n¯ır, n¯ıyir, n¯ıvir —

Ma. ni˙n-

lal ni˙n-˙nal-

To. n¨ım n¨ım-

Ko. n¯ım nim-

Kod. ni

-ga ni

-ga-

Ka. n¯ıM (n¯ım, n¯ıvu ), n¯ı˙n-ga

l nim(m)-

Tu. (dial) n¯ı-ku

l¨ı, in-ku

lu, ¯ıru ni-ku

le-, in-ku

le-, ¯ı r e -

Te.

∗

n¯ır, ¯ı r u —

Go. (dial) nim-a

t, nim-e

t; im-e

t, imm-

¯a

t —

Kui ¯ı r , n¯ım (BB) nim-

Pengo ¯ep(e

)—

Man

da ¯ı m —

Kol. n¯ır im-

Nk. n¯ır im-

Nk. (Ch.) ¯ım im-

Oll. ¯ım im-

Gad. ¯ım imm-

r. n¯ım nim-

Malt. n¯ım nim-

Early Telugu, hence Kol. Nk. n¯ıv/¯ı v , but note the oblique is the regular -in like the other

Central Dravidian languages.

As in the case of the ﬁrst person, there are two ways of forming the obliques: (a) by

shortening the long vowel of the nominative, a feature of the whole family, i.e. nin;

(b) by dropping the ﬁnal consonant of the nominative, i.e. n¯ı; the latter is an innovation

of South Dravidian II, which has both types of obliques, the (a) type in the accusative

and the (b) type in all other cases.

The vowel ¯e instead of ¯ı in Pengo in the second

person is aberrant and cannot be explained now.

The second-person-plural forms (table 6.6b) in Group IIA are derivable from PD

∗

n¯ım/

∗

nim- with the following subsequent changes: (a) addition of the common plural

∗

-ka

l to the plural form as an alternative to the basic form, perhaps as a polite expression

in South Dravidian I; (b) the regular dropping of initial n- in Central Dravidian, and in

variant forms in Tu

lu, Telugu, Kui and Pengo, as a parallel development to their sin-

gulars; (c) the mismatch between the nominative and oblique in Kolami–Naiki can be

explained only by invoking borrowing from Early Telugu in the form

∗

n¯ır. Actually

dagu of South Dravidian I is unique in independently developing the (b) type of oblique in

the ﬁrst and second person singular in dative and genitive, e.g. dat: n¯a-k¨ı ‘to me’, n¯ı-k¨ı ‘to you’

(Ebert 1996: 31–2).