NOUN CLASS AGREEMENT IN NEPALI

1. INTRODUCTION

1.1 Is Nepali a genderless language?

Priestly (1983:345) counts Nepali² as one of those Indo-Iranian languages of South Asia, which have turned to be genderless. However, this paper shows that Nepali has not only retained and developed its characteristic Indo-European gender but it has also innovated numeral classifier system similar to East Asian and Southeast Asian languages to categorize its nouns. Nepali a language which uses two sets of ‘classifiers’ like some Dravidian, Iranian and Arawak languages (Aikhenvald 2000:185).

The total set of Nepali nouns can be classified on the basis of difference in agreement patterns. However, grammatical feminine gender is limited to the human category of nouns. In the nonhuman class the feminine morpheme is lexically limited to sex distinction of big animals and diminutive meaning in the big count nouns, but it does not participate in agreement distinctions.

Pluralization and regular or irregular distribution of plural morphemes in agreement ‘targets’ are also instrumental in the subcategorization of noun classes both in the human and in the nonhuman categories. Along the nonhuman axis mass nouns cannot take plural morpheme even when more than one type is coordinated (see 34). Inanimate count nouns are also reluctant to supply plural targets even when the noun is pluralized or more than one variety of the count nouns is coordinated in the subject position (see 33). In one example of the data (32) even when the subject noun refers to more than one animates, there is an option of singular marking in the targets possibly to give generic interpretation.

The plural morpheme is grammaticalized as honorific marker along the human axis. There are four honorific grades or classes of nouns. The royal class is characteristic not to take both feminine and singular markings in the targets (see 38). The oldest honorific grade (MGH) can take both feminine and honorific markings with the feminine subject (see 35) and the second honorific grade (HGH) can take feminine marking in the NP, but nonfeminine singular in the VP (cf. 36 and 37).

In this way agreement class in Nepali is a functions of regularity and constraints of feminine and plural markings in agreement targets. Since the semantic classification of all nouns (including their honorific grades) is intertwined with the regularity and constraints on feminine and plural marking in their agreement targets, we have included even the honorific grades among the noun classes.

 

1.2 Gender determining criteria: strictly semantic

Following Corbett (1991:8) and Corbett and Fraser (2000) gender assignment in Nepali is ‘strictly semantic’.

Corbett (2006:155-156) correlates variable agreement with ‘semantic agreement’ evidently from the speaker’s point of view. In Nepali feminine morphemes in the noun cannot trigger feminine morphemes in the targets provided the referent noun is human (cf. 12b, 13b, 14b); therefore in order to control agreement targets the speaker should know whether the noun in question is human or nonhuman. On the other hand there are many feminine nouns that do not have feminine suffixes attached to them (see 4), but their agreement targets are decided by agreement class assigned by the speaker. However, in a subset of data we will find both feminine (and corresponding nonfeminine) and plural (and corresponding singular) morphemes regularly used in the targets (see 31).

As a typical agreement target and the shape of the triggered morpheme can be decided by the lexical (and ‘online’ pragmatic) meanings of the mass nouns, so is the case of assigning plural morphemes (or not) in the targets of subclasses within the range of both human and nonhuman classes (cf. 32, 33, 34). Since an honorific grade is pragmatic, deictic or contextually assigned ‘on-line’ and not strictly lexical, the dictionary meaning of the noun also does not help to determine a particular agreement class of the subject noun (cf. 31b, 35b, 37 and 38).

Even within the inanimate count nouns, it is the speaker who decides whether the multiplex of referents are to be visualized as unbounded and generic in which case the individual boundaries between the categories are ignored and visualized fuzzy. When the boundary is ignored the multiplex of the referents typically triggers singular morpheme in the targets (cf. 32, 33 and 34).

In such a situation of variable gender of a single noun since the morphological shape of a noun does not always help to ascertain which particular agreement class the noun belongs to, it is the agreement patterns the noun takes, that specifies its gender or noun class even for the native speaker or the native listener, and therefore we can say that the assignment of gender or noun class or agreement class in Nepali is semantic for the speaker and syntactic for the listener. This strategy strictly applies even to problematic nouns which are characterized by variable genders and variable honorific classes. Even when the gender of a noun is morphologically ‘overt’ or ‘covert’ (Corbett 1991:62) sex and honorific semantic variables of the noun are determined by the patterns of agreement they take with its targets.

 

1.3 Terminology: gender, noun class or agreement class

According to Corbett (1991:146) gender and noun class are linguistically synonymous; they are different only due to their traditions; Dravidian tradition uses the word ‘gender’ to denote the same category of nouns which the East Caucasian tradition likes to call ‘noun class.’ Corbett has reserved the term ‘gender’ for the distinction of animacy and sex. We have also reserved the term ‘gender’ for the linguistic distinction of sex to respect traditional grammars of Nepali and in order to include other varieties of nominal agreement types (like honorific, etc.) we have chosen the term ‘noun classes’ in this paper. However, in a broad sense the terms‘gender’,‘ noun classes and ‘agreement class’ are used synonymously.

 

1.4 Organization of the paper

Nepali agreement class of nouns is a function of number and gender (sex) marking affixes in the modifiers (adjective, possessive, classifier, past participle and genitive postposition), verb inflections and pronoun substitution. Corbett (1991:151) calls gender variation in these words as ‘target genders’ corresponding to variations in noun, which are termed as ‘controller genders’. Section 2 deals with the feminine morphology of targets and controllers. Section 3 gives the detailed introduction to patterns of nominal agreement. Section 4 deals with the grammaticalization of items which interact in determining noun classes in the language. Section 5 is allotted to the sociolinguistic processes that are instrumental in gender loss, gender shift and the emergence and decay of honorific classes. Section 6 raises issues and conclusions.

2. feminine vs. nonfeminine

Feminine is grammatically marked in Nepali. It can regularly be defined in the target genders although it is not always easy to infer feminine gender in the nominal morphology. A large number of nonfeminine words in Nepali end in <o> which could clearly be labeled as ‘masculine’, but the morphology of the rest of the nonfeminine words (nouns and their modifiers) does not give any clue about the masculine gender in contrast with the morphology of the feminine. That is the reason why we have chosen the dichotomy feminine vs. ‘nonfeminine’ rather than ‘masculine’ vs. ‘nonmasculine’. In addition to this clear morphological distinction of <o> vs. <i> there are both male and female denoting nouns without these endings and their gender is semantically determined by native speakers although even for the native speaker as a listener the exact agreement class is difficult to predict. Besides, masculine is default or unmarked, therefore the label ‘nonfeminine’ is chosen.

 

2.1 Feminine morphology of nouns

Feminine and nonfeminine contrast is characteristic of Nepali nominal morphology. However, there is a mismatch between the lexical feminine and syntactic feminine, since as we will see in Section 3, Nepali disallows feminine agreement in the nonhuman class of nouns, lexical feminine of nonhuman animates only denotes distinction in sex while the same morpheme in inanimates is used to contrast diminutive from augmentative (see Section 4). The disability of nonhuman controllers to have any control of feminine marking over the possible targets puts the nonhuman animate class to realize no gender distinction of sex and similar behavior is shown by royal subclass of the human class.

Human nouns in Nepali are either feminine or nonfeminine. Majority of the words referring to females in the language end with <(in)i> while the majority of the female given names end in <a> in addition to a few others that end in <i>.

 

(1) Female given names

a. krishn-a, radh-a, sit-a, shyam-a, gom-a, kʌmʌl-a, rambh-a, bidy-a, sharʌd-a, etc.

b. sʌrʌswʌt-i, lʌksm-i, kirt-i, srut-i, riddh-i, siddh-i, nan-i, srisʈ-i, sant-i, kant-i, etc.

(2) Nouns referring to females

keʈ-i ‘girl’, tshor-i ‘daughter’, aima-i ‘woman’, deb-i ‘goddess’, swasn-i ‘wife’, kak-i ‘paternal aunt’, bʌɦin-i ‘younger sister’, did-i ‘elder sister’,

(3) Nouns meaning ‘wife of’

kʌmi-ni ‘wife of Kami’, sʌrki-ni ‘wife of Sarki’, kʌrki-ni ‘wife of Karki’, sahu-ni ‘wife of Sahu’, (exception nati-ni ‘granddaughter’)

 

In Nepal most of the given names come from Sanskrit where the majority of female referring nouns end with <a> or <i>. In Nepali it is remarkable that except for a few words like < nati-ni> ‘granddaughter’ and <sad̤u-ni> ‘female hermit’ the suffix <(i)ni> denotes wife relation while <a> or <i> simply denote the referents of female sex.

There are, however, apparently female referring nouns (mostly kinship terms) which do not end in <a>, <i> or <ini>. Feminine gender of such words is ‘covert’ rather than ‘overt’. Gender in such words should be specified in the lexical entry. To complicate the issue, there are some ‘defective nouns’, ‘hybrid nouns’ and ‘double-and multiple-gender nouns’ (Corbett 1991:175-183) which typically take multiple agreements although morphologically there is no way to assess their genders.

 

(4) Formally irregular nouns of feminine gender

ama ‘mother’, phupu ‘paternal aunt’, maidzu ‘maternal aunt’, nʌndʌ ‘hustand’s younger    sister’, amadzu ‘husband’s elder sister’, sasu ‘mother-in-law’

 

In several Western dialects (like Jumli and Baitadeli) of Nepali the nouns with the feminine suffix show agreement with feminine targets, but in the standard Nepali the feminine marking suffixes in the nonhuman category of nouns have lost gender distinction through agreement contrasts.

 

2.2 Feminine morphology in targets (modifiers and predicate)

Allomorphs of feminine morpheme are neutralized as <i> in the targets (possessives, adjectives, classifiers, participles and verbs. Only in a single feminine verb <tsh-e> ‘is’ (3.SG.NP-F) the feminine morpheme <i> has phonological lowering as <e>.

 

(5) Possessives: me-r-i (I-POS-F), ham-r-i (we-POS-F), te-r-i (you.LGH-POS-F), tim-r-i (you.MGH-F)

(6)   Postposition: k-i (POS-F)

(7) Adjectives: kal-i (black-F), kantsh-i (youngest-F), ramr-i (elegant-F), san-i (small-F)

(8) Classifiers: eu-ʈ-i (one-CL-F), dui-ʈ-i (two-CL-F), tin-ʈ-i (three-CL-F), oʈ-i (CL-F)

(9) (Past) Participles: lekʰ-e-k-i (write-P-POS-F) ‘written (F)’, bʌs-e-k-i (sit-P-POS-F) ‘sat’ (F), gʌ-e-k-i (go-P-POS-F) ‘gone’ (F)

(10) Verbs: lekʰ -i (write-F) ‘wrote’ (F), bʌs- i (sit-F) ‘sat’ (F), gʌ-i (go-F) ‘went’ (F)

(11) MGH verbs: (a). Past: lekʰ -i-n (write-F-PL) ‘wrote’ (F.MGH.SG), bʌs- i-n (sit-F-PL) ‘sat’ (F.MGH.SG), gʌ -i-n (go-F-PL) ‘went’ (F.MGH.SG) (b). Nonpast: lekʰ -tsh-i-n (write-NP-F-PL) ‘writes’ (F.MGH.SG), bʌs- tsh-i-n (sit-NP-F-PL) ‘sits’ (F.MGH.SG), dza-n – tsh-i-n (go-NP-F-PL) ‘goes’ (F.MGH.SG)

3. agreement patterns

Gender is the classification of nouns in a language in terms of the varieties of agreements nouns take with other words within a noun phrase or a clause or across clause boundaries. However, as we have noted in the introduction that noun class or gender assignment in Nepali is strictly semantic and also that the morphological shape of a noun does not always help to ascertain which particular agreement class the noun belongs to. This strategy strictly applies even to problematic nouns which are characterized by variable genders and variable honorific classes.

 

3.1 Gender in the clause domain: Subject-verb agreement

The verb form shows the distinction between HUMAN and NONHUMAN. The DEFAULT HUMAN (LGH) and the FIRST HONORIFIC (MGH) classes show FEMININE and NONFEMININE contrast. The SECOND HONORIFIC (HGH) and the THIRD HONORIFIC (royal) do not contrast for feminine and plural in the verb. They use the 3^rd person singular nonfeminine or default class marking in the verb, but they have their separate distinct verb structures. These structures predict particular honorific classes of the subject noun.

 

(12a) keʈ-o                  bʌs-y-o (nonfeminine)

child-NF.SG      sit-P-NF.SG

‘The boy sat down’

(12b). keʈ-i                   bʌs-i (feminine)

child-F             sit-F

‘The girl sat down’

 

This pair shows the subject verb agreement pattern of the default (LGH) class. These examples clearly show that Nepali has feminine and nonfeminine (masculine) contrast in agreement. These two separate patterns in agreement mean that <keʈ-o> ‘boy’ and <keʈ-i> ‘girl’ belong to different genders which gives the impression that <o> in the subject noun triggers <o> in verb and so does the <i>  marking in the noun.

However, this contrast of agreement between feminine and nonfeminine (masculine) genders is limited to human class of nouns. In the nonhuman class there is no such contrast.

 

(13a). batsh-o bʌs-y-o

calf-M   sit-P-NF

‘The male calf sat down’

(13b). batsh-i   bʌs-y-o

calf-F   sit-P-NF

‘The female calf sat down’

(14a). ɖal-o                              laɖ-y-o

bamboo.basket-NF      fall-P-NF

‘A big bamboo basket fell down’

(14b). ɖal-i                         laɖ-y-o

bamboo.basket-F         fall-P-NF

‘A small bamboo basket fell down’.

 

In this set of examples each of the nouns of nonhuman category in (13b) and (14b) is marked feminine (F), however, the verbs are marked nonfeminine (NF). Therefore all nouns in (13) and (14) belong to the same gender for the lack of agreement contrasts. If we compare this set of examples with (12b), both the nouns and verbs are marked feminine (F).

The examples show that feminine gender is sensitive only if the referent noun is human. Thus, feminine agreement is a test that classifies nouns primarily into human and honhuman classes.

 

3.2 Gender in the NP domain: modifier-noun agreement

In the NP domain the general classifier for COUNT nouns <oʈ-o/a> shows primary contrast between MASS vs. COUNT nouns. From among the COUNT nouns, it further shows the agreement contrast between FEMININE and NONFEMININE. Rest of the modifiers in the NP domain primarily shows agreement contrast between FEMININE and NONFEMININE where the NONFEMININE shows number agreements.

 

A.  Possessor-noun agreement

 

(15a).   mer-o               keʈ-o

my-NF             child-NF

‘My boy’

(15b).   mer-i    keʈ-i

my-F    child -F

‘My girl’

(16a).   mer-o               batsh-o

my-NF             calf-NF

‘My male calf

(16b).   mer-o               batsh-i

my-NF             calf-F

‘My female calf’

(17a).   mer-o               ɖal-o

my-NF             bamboo.basket-NF

‘My big bamboo basket’

(17b).   mer-o               ɖal-i

my-NF             bamboo.basket-F

‘My small bamboo basket’

 

B.  Adjective-noun agreement

 

(18a).   kal-o                keʈ-o

black-NF         child -NF

‘Black boy’

(18b).   kal-i                 keʈ-i

black-F            child -F

‘Black girl’

(19a).   kal-o                batsh-o

black-NF         calf-NF

‘Black male calf

(19b).   kal-o                batsh-i

black-NF         calf-F

‘Black female calf’

(20a).   kal-o                ɖal-o

black-NF         bamboo.basket-NF

‘Big black bamboo basket’

(20b).   kal-o                ɖal-i

black-NF         bamboo.basket-F

‘Small black bamboo basket’

 

 

 

C. Classifier-noun agreement

(21a).   eu.ʈ-o/a            keʈ-o

One.CL-NF     child-NF

‘A boy’

(21b). euʈ-i                 keʈ-i

one-F               child -F

‘A girl’

(22a).   eu.ʈ-o/a batsh-o

one.CL-NF calf-NF

‘A male calf

(22b).   euʈ-o/a             batsh-i

one-NF            calf-F

‘A female calf’

(23a).   euʈ-o/a             ɖal-o

one-NF            bamboo.basket-NF

‘A big bamboo basket’

(23b).   euʈ-o/a             ɖal-i

one-NF            bamboo.basket-F

‘A small black bamboo basket’

 

D. Postposition-noun agreement

 

(24a)   tshor-i                          k-o                   tshor-o

son-F                           of-NF.SG        son-NF.SG

‘daughter’s son’

(24b)   tshor-i                          k-i                    tshor-i

son-F                           of-F                 son-F

‘daughter’s daughter’

 

E. Past participle-noun agreement

 

(25a)    roj-e-k-o                                  keʈ-o

choose-P-GEN-NF.SG           child -NF.SG

‘chosen or selected boy’

(25b)    roj-e-k-i                                   keʈ-i

choose-P-GEN-F.SG              child -F

‘chosen or selected girl’

 

Both (D) and (E) can be taken as subsets of (A).

 

3.3 Gender agreement across clause boundaries: ‘pronominal genders’

Hockett (1958:231-233) defines genders as ‘classes of nouns reflected in the behavior of associated words’ and also tries to establish genders of nouns in English by pronoun substitution method. Corbett (1991:5) has reluctantly labeled such a system of gender as ‘pronominal gender systems’. It may be because pronouns and pronominal expressions do not co-occur with their antecedents or controllers in the same clause or ‘governing category’ (Chomsky 1981:188) while other ‘controllers and targets’ (Corbett 1991: Section 6.3) usually co-occur in the same clause or noun phrase.

Lexical gender of relative, interrogative and indefinite pronouns and pronominal words in Nepali clearly distinguish HUMAN vs. NONHUMAN nouns. Feminine and nonfeminine contrast is not a characteristic of Nepali pronouns.

 

(26) Interrogative, indefinite and relative pronouns:

a. Interrogative pronouns: ko ‘who?’, ke ‘what?’

b. Indefinite pronouns: ko-hi ‘someone’, ke-hi ‘something’

c. Relative pronouns: dzo ‘who’, dze ‘what’

 

This phenomenon of human and nonhuman contrast is found even in the 3^rd person pronoun(s).

 

(27) Personal pronouns:

a. u ‘he/she’, uni hʌru ‘they’ (Personal pronoun human)

b. tyo ‘he/she/it/that’, tini hʌru ‘they’ (Personal and demonstrative human and nonhuman)

 

The difference between noun and pronouns lies in the fact that pronouns never co-occur with their controller nouns in the same clause.

If we merge the clause domain (3.1) and all the five subtypes of NP domain (3.2), we will come up with the following patterns where:

1. ‘Pronominal genders’ as established by contrasting agreements between antecedent nouns and relative, interrogative and personal anaphora identify only human and nonhuman categories of nouns. Gender loss in pronouns like this is against the expectation of Priestly (1983:340) who believes that gender loss in pronouns should be the last event after nouns followed by adjectives.
2. Although in the lexical level feminine and nonfeminine contrast may be found in both the human and nonhuman classes of noun, only nouns of the human class can trigger feminine agreement in the targets (verbs, modifiers, postpositions and participles).

 

(28a) ko    bʌs-y-o?                 

        who  sit-P-NF.SG

‘Who sat down?’

(28b) keʈ-o                   bʌs-y-o.            u          bʌs-y-o.            dzo                   bʌs-y-o,            u/ tyo…

       child -NF.SG         sit-P-NF.SG      he        sit-P-NF.SG      who.REL         sit-P-NF           he/that

‘A boy sat down’.                         ‘He sat down’              ‘He who sat down…’

(28c) keʈ-i        bʌs-i.                u          bʌs-i.                dzo                   bʌs-i,    u/ tyo…

       child -F.SG                        sit-P-F.SG        she       sit-P-F.SG        who.REL         sit-P-F he/that

‘A boy sat down’.                         ‘He sat down’              ‘She who sat down…’

(29a) ke    lʌɖ-y-o?           

        what  fall-P-NF.SG

‘What fell down?’

(29b) batsh-o               lʌɖ-y-o.            tyo       lʌɖ-y-o.                        dze       lʌɖ-y-o, tyo…

        calf-NF.SG          fall-P-NF.SG    that      fall-P-NF.SG    which   fall-P-NF, that..

‘A he calf fell down’.                               ‘It fell down’   ‘The thing which fell down…’

(29c) batsh-i                lʌɖ-y-o.            tyo       lʌɖ-y-o.                        dze       lʌɖ-y-o, tyo…

        calf-F.SG fall-P-NF.SG    that      fall-P-NF.SG    which   fall-P-NF, that..

‘A she calf fell down’.                  ‘It fell down’   ‘The thing which fell down…’

 

The examples in (28) show that only human nouns serve as antecedents of interrogative, relative and personal pronouns <ko> ‘who?’, <dzo> ‘who’ (REL), <u> ‘he/ she’ (HUM) while those of (29) show nonhuman nouns serving as antecedents of interrogative, relative and personal pronouns <ke> ‘what?’, <dze> ‘which’ (NHUM.REL) and <tyo> ‘that’ (DPR). There is a special 3^rd person pronoun <u> for human, but this is not the case for nonhuman class. The demonstrative pronoun <tyo> can be used for both human and nonhuman categories equally.

An interesting thing in this ‘pronominal gender’ system is that even the human baby is mapped by the nonhuman anaphora.

 

(30a) tshor-i                 le         ke         pa-i?

offspring-F        ERG     what     get-F.P

‘What did the daughter beget, son or daughter?’

(30b) tshor-i                le         tshor-o                         pa-i.

         offspring-F         ERG     son-NF.SG       get-F.P

‘The daughter begot a son’

(30c) tshor-i                 le         thor-i   pa-i.

          offspring-F        ERG     son-F   get-F.P

‘The daughter begot a son’

(30d) thor-i      le         je                     pa-i,     buhari                   le    tyʌhi                 pa-i.

offspring-F        ERG     REL.NHUM     get-F.P daughter-in-law   ERG  that.EMP        get-F.P

‘The daughter-in-law birth to a child of the same sex as the daughter gave’.

 

3.4 Interaction of number and sex categories in agreement patterns

Nonhonorific or Low Grade Honorific (LGH) category is the default noun class in Nepali, because this class of noun regularly agrees with all targets in terms of both number and gender. If we compare (31a) and (31b), we can see that in every target (and controller noun) there is a change in suffix from <o> to <i> to denote change from nonfeminine (masculine) to feminine. Similarly, if we compare (31a) and (31c), we find that for every suffix <o> in (31a) there is <a> in the modifiers and <e>³ in the verb. The kind of regular contrast in gender and number that we witness as a characteristic of this default or basic level class of noun gradually disappears with nouns of both the human and nonhuman axes. The higher up we go along the human axis from the default class, we enter a range of honorific subgenders where gradually the relevance of feminine morpheme diminishes and the relevance of plural morpheme increases.

On the other hand, if we move down along the nonhuman class, the relevance of feminine morpheme gradually disappears from the target and so does the relevance of plural morpheme. This kind of reluctance of feminine and plural morphemes interacting with different classes of noun is what has increased the patterns of nominal agreement in Nepali.

 

(31a) mer-o                 eu.ʈ-o/a                        kal-o                keʈ-o                bʌs-y-o

my-NF.SG           one.CL-NF.SG           black-NF.SG   child-NF.SG   sit-P-NF.SG

‘My black boy sat down’.

(31b) mer-i                  eu.ʈ-i                kal-i                 keʈ-i                 bʌs-i

my-F                     one.CL-F         black-F            child-F                         sit-P-F

‘My black girl sat down’.

(31c) mer-a                 dui-ʈ-a              kal-a                keʈ-a                bʌs-e

my-PL                  two-CL-PL     black-PL          child-PL          sit-P.PL

‘My two black boys sat down’.

 

This kind of similar marking (cf. 31b) across all the targets and controllers is termed ‘alliterative concord’ by Corbett (1991:117).

 

3.5 Plural morpheme as a determinant of subgenders in the nonhuman class

3.5.1 Animate vs. default (LGH) human class

The agreement pattern of nonhuman animate class (32) is very similar to the default human class (LGH) exemplified by (31) with respect to the plural morpheme. They are separated as different classes only by the presence (31b) vs. absence (32b) of feminine markers in the targets.

Another marked agreement contrast between the animate and the default (LGH) human class is characterized by the possibility of using Singular-for-Plural morpheme in the targets of animate class (32c), which is impossible with the default human class.

This morpho-semantic feature is apparently spread from the mass nouns where countability is not relevant. Nepali speakers sometimes treat the mass of nonhuman objects as ‘unbounded’ (Talmy 2000:50). On such a perspective even the ‘multiplex’ of ‘discrete’ count nouns often agree with singular targets (see 32c, 33b). Such a cognitive representation is facilitated by the extension of the singular morpheme for more than one referent. Priestly (1983) has given the name ‘collectivity’ to such a feature of multiple agreement of inanimate nouns in Indo-European.

 

(32a) mer-o                 eu.ʈ-o/a                        kal-o                batsh-o            lʌɖ-y-o

my-NF.SG           one.CL-NF.SG           black-NF.SG   calf-NF.SG     fall-P-NF.SG

‘My black he-calf fell down’.

(32b) mer-a                 dui-ʈ-a                          kal-a                batsh-a            lʌɖ-e

my-NF.SG           two.CL-NF.SG           black-NF.SG   calf-NF.SG     fall-P-NF.SG

‘My two black calves fell down’.

(32c) mer-o                 pʌtsas oʈ-a                   mʌɦʌg-o                      bʌstub̤au          mʌr-y-o

my-NF.SG           50       CL-NF.SG       valuable-NF.SG          livestock          fall-P-NF.SG

‘My fifty black valuable calves fell down’.

(33a) mer-a                 dui-ʈ-a                          kal-a                ɖal-a                lʌɖ-e

my-NF.SG           one.CL-NF.SG           black-NF.SG   basket-PL        fall-P.PL

‘My two black baskets fell down’.

(33b) mer-o                 dui-ʈ-a                          kal-o                ɖal-o                lʌɖ-y-o

my-NF.SG           one.CL-NF.SG           black-NF.SG   basket-NF.SG             fall-P-NF.SG

‘My two black baskets fell down’.

(34a) mer-o                 *eu.ʈ-a                         pani                 kʰʌs-y-o

      my-NF.SG                        one.CL- NF.SG          water               fall-P-NF.SG

‘My water fell down’

(34b) mer-o                 dal,      b̤at,      tʌrkari pak-yo.

      my-NF.SG                        lentil    rice      curry    cook-P-NF.SG

‘My rice, lentil soup and vegetables are readily cooked’

 

3.5.2 Mass (unbounded) vs. count (bounded) nonhuman

According to Talmy (2000), mass nouns are ‘unbounded’ and ‘continuous’ while count nouns are ‘bounded’ and ‘discrete’. Such semantic features of the mass nouns preclude them to associate with counting words and classifiers for count nouns like <oʈ-a> on the one hand and with plural marking targets on the other (see 34).

This characteristic feature of mass nouns contrasts them with other inanimate and animate count nouns, because count nouns can be pluralized and so can associate and agree with plural morphemes and classifiers for the count class.

 

3.6 Plural and feminine morphemes as determinants of subgenders in the honorific

The class of human nouns crosscuts between sex (gender) and honorific social deixis. Within this class feminine and plural morphemes interact as determiners of subgenders or subclasses.

 

3.6.1 First grade honorific (MGH)

This is the first historical honorific in Indo-European and is thus shared by the majority of sister languages. The typical structural definition of this honorific grade is Plural-for-Singular morpheme. It means in this subclass of human nouns there is no singular morpheme: plural morpheme is used for both controllers and targets even to denote singular referents.

 

(31a) mer-o                 eu.ʈ-o/a                        kal-o                keʈ-o                bʌs-y-o

my-NF.SG           one.CL-NF.SG           black-NF.SG   child-NF.SG   sit-P-NF.SG

‘My black boy sat down’.

(31c) mer-a                 dui-ʈ-a              kal-a                keʈ-a                bʌs-e

my-PL                  two-CL-PL     black-PL          child-PL          sit-P.PL

‘My two black boys sat down’.

(35a) mer-a                eu.ʈ-a               kal-a                keʈ-a                bʌs-e

my-PL                  one-CL-PL      black-PL          child-PL          sit-P.PL

‘My black boy (MGH) sat down’.

 

If we compare (35a) with (31a), we will see that although in both the sentences the referent noun is singular, but in (35a) the noun and all its targets are marked with the plural suffix <a>, because in (35a) the noun is in the first honorific (MGH) grade. On the other hand if we compare (35a) with (31c) we will see that plural markers are shared by both. It also implies that number is neutralized as plural in the first honorific class of nouns.

Another remarkable feature of this honorific grade is the presence of plural morpheme in both the feminine and nonfeminine verbs. Besides, this is the only honorific grade where the verb is marked plural. Plural morpheme is not found in the verbs of special honorific grades higher than this. We can compare the feminine noun in the first honorific grade (MGH) (35b) with the feminine noun in the default (LGH) grade (31b).

 

(31b) mer-i                  eu.ʈ-i                kal-i                 keʈ-i                 bʌs-i

my-F                     one.CL-F         black-F            child-F                         sit-P.F

‘My black girl sat down’.

(35b) mer-i                  eu.ʈ-i                kal-i                 keʈ-i                 bʌs-i-n

my-F                     one.CL-F         black-F            child-F                         sit-P.F-n

‘My black girl (MGH) sat down’.

 

We will also see that there is no corresponding plural morpheme for feminine in the NP domain; it is only found in the clause domain of this honorific grade only.

 

 

3.6.2 Second honorific grade (HGH)

This honorific grade is morphologically similar to the first honorific grade (MGH) in the NP targets, but there are neither feminine nor plural morphemes in the VP, characteristics of both the feminine and nonfeminine nouns of the first honorific grade.

 

(35a) mer-a                 eu.ʈ-a               kal-a                keʈ-a                bʌs-e

my-PL                  one-CL-PL      black-PL          child-PL          sit-P.PL

‘My black boy (MGH) sat down’.

(36) mer-a                   eu.ʈ-a               kal-a                kak-a               bʌs-nu              b̤ʌ-y-o

my-PL                 one-CL-PL      black-PL          uncle-PL          sit-NOM          be-P-NF.SG

‘My black uncle (HGH) sat down’

 

The lack of sex and number distinction in VP of both the singular and the plural contexts is characteristic of Special (HGH and royal) honorific (2^nd and 3^rd) grades.

3.6.3 Third honorific (royal) grade

The 3^rd honorific grade (royal) shares the following features with the 2^nd honorific grade (HGH):

(a)    Use of plural morphemes for nonfeminine singular nouns and their targets in the NP and the neutralization of number with the plural morpheme.

(b)   Use of nonfeminine singular form in the VP

Following are the contrasting features of the 3^rd honorific against the 2^nd grade:

(a)    The 2^nd grade maintains feminine marking in the targets of the NP domain while the 3^rd grade does not use feminine morpheme in any of the targets.

(b)   Special compound verb and light verb structures and word substitution exemplified by (40) and (41) are characteristics of the 3^rd honorific grade. Feminine morpheme may be limited to the nouns that control target genders.

 

(37a) hamr-i                kantsh-i            kak-i                bʌs-nu              b̤ʌ-y-o

our-F                    youngest-F      uncle-F            sit-NOM          be-P-NF.SG

‘Our youngest aunt (HGH) sat down’.

(37b) hamr-i                kantsh-i            kak-i                le         her-nu              b̤ʌ-y-o

our-F                    youngest-F      uncle-F            ERG    see-NOM        be-P-NF.SG

‘Our youngest aunt (HGH) saw (something)’.

(38a) hamr-a               kantsh-a           shahadzyad-i   raj        b̤ʌ-y-o

our-PL                  youngest-PL    prince-F           reign    be-P-NF.SG

‘Our youngest princess (Royal) sat down’

(38b) hamr-a               kantsh-a           shahadzyad-i   baʈʌ     nʌdzʌr              b̤ʌ-y-o

our-PL                  youngest-PL    prince-F           ABL    eye.PERSIAN                        be-P-NF.SG

‘Our youngest princess (Royal) saw (something).

 

The following formal features related to deviations in gender and number, are shared by these

Special honorific grades:

a. For the nonfeminine referents they have maintained the Plural-for-singular strategy of the original (MGH) honorific grade, but this strategy is limited only to the noun phrase targets and not to the verb phrase ones.

b. Since the feminine lacks the corresponding plural forms both in the controller nouns and their targets, there is no honorific distinction in the noun phrase for the first new honorific (HGH), but the second new honorific (Royal) grade has lost gender distinction but taken the Plural-for-singular strategy of the nonfeminine gender. This is one of the characteristic formal distinctions between the two special grades of the new honorific.

c. The specialty that takes these new honorifics distinct from the original one is their absence of both feminine and plural morphemes in the verb phrase.

 

3.6.4 Royal vs. mass nouns

Mass and royal nouns are the two extremes in the range of noun classes taking the nonhonorific nonfeminine default noun as the base. The range of feminine (vs. nonfeminine) gender contrast starts from the base and goes high upstream along the human axis, but the farther we go from the base the lesser is the effect of feminine marking. As we reach the highest extreme (royal class), such a contrast of sex-gender is limited only to the morphology of lexical nouns and not to the targets.

The default noun class is the base even for the plural marking. Plural marking is more relevant downwards along the nonhuman axis, but even here the relevance of pluralization goes on decreasing as we go farther from the base. Pluralization is irrelevant in both the extremes (royal and mass nouns). However, there is only plural morpheme to indicate uniplexity of the referent noun on the highest category (royal) of human class, but there is only singular morpheme in the targets used to denote multiplexity of the referent mass noun. The function of plural morpheme as honorific starts from the first honorific (MGH) grade upwards, but its effect ends in the VP abruptly above this point although its effect continues up to the highest extreme in the NP.

 

3.7 Semantic gender resolution rules

3.7.1 Person resolution

a. When (pro)nouns of two different persons are coordinated, the higher one controls resolution.

 

(39) tʌ̃                  rʌ      mʌ gʌ-y-ʌũ

you.LGH    and      I   go-P-1PL

‘You and I went’

(40) tʌ̃                rʌ      u      gʌ-y-ʌu

      you.LGH   and  (s)he   go-P-2PL

‘You and he went’

 

3.7.2 Number resolution

When nouns of different number are coordinated in the subject position, the resolved number is plural. The examples (39) and (40) can be taken as examples.

 

3.7.3 Gender (sex) resolution

When nouns of different gender (sex) are coordinated, naturally, the nonfeminine one controls agreement, because there is no feminine plural form in the language.

 

 

 

3.8 Syntactic and mixed gender resolution rules

When nouns of two honorific grades are coordinated, if the noun which is closer to the verb is higher, it controls verb agreement (41a). However, if the lower noun is closer to the verb (41b and 41c), there is often a conflict, in which case the higher one wins as an alternative. This is an example of mixed resolution.

 

(41a) b̤ai                  rʌ      ama        au-nu           b̤ʌ-y-o

younger.brother   and    mother   come-NOM  be-P-3.NF.SG (HGH)

(41b) ama        rʌ       b̤ai             a-e

mother   and     y.brother    come-3.PL.P (LGH)

(41c) ama rʌ b̤ai au-nu b̤ʌ-y-o (HGH)

‘My mother and younger brother came’

 

3.9 Number of agreement classes

Combination of contrasts in agreement nouns show with different agreement targets (verbs, pronouns, classifer and other modifiers) result in ten noun class matrices. The primary division is between HUMAN and NONHUMAN. The HUMAN class is subdivided into DEFAULT and HONORIFIC. The HONORIFIC class is subdivided into ORIGINAL and SPECIAL. The SPECIAL honorific splits into ROYAL and NONROYAL. Except for the ROYAL class which does not contrast for sex-gender, each of the other HUMAN classes splits into FEMININE and NONFEMININE.

The NONHUMAN class primarily subdivides into MASS vs. COUNT. The COUNT noun splits into ANIMATE and INANIMATE. The INANIMATE class splits into BOUNDED AND UNBOUNDED.

 

Figure : Agreement classes

 

4. semantics

The agreement class of nouns in Nepali is determined by the interplay of gender (SEX) and number (QUANTITY) marking affixes. These two categories of grammar are assisted by case marking postpositions (cf. 37 and 38) and pronoun substitution (see 30) in syntax.

 

4.1 Gender

SEX is a semantic category of ANIMATE nouns. Among SEX, QUANTITY, and HONOR it is an inherent property of the ANIMATE objects. QUANTITY is a secondary feature of an object and is assigned by context. HONOR is a tertiary feature of HUMAN nouns assigned by pragmatic social deixis.

 

4.2 Prototype extension of feminine

Gender distinction is a characteristic of an Indo-European language like Nepali. In the Nepali speaking society sex distinction is culturally relevant. It may be one of the reasons why Nepali has grammaticalized sex marking gender, but there have been prototype extensions.

a. Gender agreement is limited to HUMAN nouns. It means, culturally sex distinction is more pertinent in the human society. The full fledged agreement with all the targets of nonhonorific human class gives a hint that gender prototype is central to this level of human class.

b. Gender marking contrast is limited to lexical level of ANIMATE, INANIMATE and ROYAL classes but not reflected in agreement patterns. Since gender is not grammatical in these classes of noun, sex distinction of these classes is socioculturally less relevant for ordinary Nepali speakers.

c. The function of the feminine morpheme <i> in the inanimate class is to give diminutive (vs. augmentative) meaning. This fact simply hints to the prototype extension, because the size of females in general is smaller in nature. MALE IS BIG, FEMALE IS SMALL metaphor may have been instrumental in this prototype extension.

d. The metaphor FEMALE IS ELEGANT may have been functional in the use of feminine morpheme after the inanimate words for beautification. In songs, words like <bihan> ‘morning’ and <pʌrel-o> ‘eyelash’ get feminine morphemes as <bihan-i>  and <pʌrel-i>.

e. Generally males are slim and females are plump in Nepal. That may be the reason why the same feminine morpheme <i> that indicates short size in (c) is found suffixed with the word <ɖor-i> ‘rope’ against <ɖor-o> ‘thread’. MALE IS SLIM, FEMALE IS PLUMP metaphor may have worked for this prototype extension.

f. In the Nepali speaking society and culture leadership is associated with males. The metaphor LEADER IS MASCULINE may have been active in the suffixation of the masculine (nonfeminine) morpheme after the word <ran-o> ‘queen bee’ vs. <ran-i> ‘queen’. Leadership may also have the implication of power.

g. The other feminine morpheme <ini> which shows WIFE relation in the derivation of HUMAN nouns indicates CONTAINER or ABODE relation among some inanimate nouns. This fact seems to establish metaphorical equation between WIFE and CONTAINER.

h. Plural morpheme is grammaticalized as honorific in the nonfeminine class, therefore a nonfeminine (or masculine) noun has honorific vs. nonhonorific (default) contrast, but except for the 1^st honorific (MGH) verb form (see 35) , feminine nouns and their targets do not contrast for honorific. This absence in the honorific for feminine class may hint less prestige given to females of Nepali speaking society and culture.

i. Except for a single example <ran-o> ‘queen bee’ where semantically the masculine (nonfeminine) form is derived from feminine, the direction of derivation is always from nonfeminine to feminine. This may also imply the dependent roles women are given by culture.

j. Another point worth mentioning is that love and affection are often associated with exchange of gender morphemes. Parents often use the feminine morpheme to address their sons and the masculine morpheme to address their daughters. Parents often use (31a) and (35a) to refer to their daughter and (31b) and (35b) to refer to their son out of affection when they wish to express love or affection.

 

4.3 Number

Number is basically a grammaticalization of QUANTITY, but this category interacts with different semantic classes of noun differently. While discussing the schematic category disposition of quantity Talmy (2000:59) classifies matter around us into CONTINUOUS and DISCRETE. Continuous matter forms a MASS noun like water which does not have ‘inherent segmentation’ (Evans and Green 2006:515) or is homogeneous in contrast with a discrete matter like a tree which ‘is conceptualized as having breaks or interruptions through its composition’ (Talmy 2000:55). ‘State of boundedness’ is another configurational variable of matter. ‘When a quantity is understood as bounded, it is conceived to be demarcated’ (Talmy 2000:50). A MASS noun can be made a COUNT noun by supplying boundary like that of a pond or a river. In Nepali there are many classifiers (see Pokharel 1997) which supply boundaries for counting different shapes and sizes of bounded mass nouns similar to Tzeltal Maya (see Berlin 1968).

 

4.3.1 Reluctance for plural morpheme

Even the boundary of an inherently BOUNDED noun can be cognitively ignored in which case a bounded noun is treated as unbounded. If we look a plate from very close, its boundary disappears. Similarly the boundaries of trees disappear if we try to look at them far in the forest. In the latter case tree becomes GENERIC. The processes of supplying and ignoring boundaries seem to be instrumental in assigning singular morpheme to more than one NONHUMAN COUNT nouns in Nepali.

Sanches and Slobin (1973) have established a correlation between numeral classifiers and the use of plural marking; in that if a language uses numeral classifiers, obligatory plural marking is either absent or infrequent. On the other hand, in support of the Whorfian linguistic relativity hypothesis with cognitive experiments, Lucy (1992:73) presumes that in a numeral classifier language like Yucatec Maya, lexical nouns are not specified for units; so pluralization is not obligatory, but unitization by supplying obligatory classifiers is obligatory in contrast with a non-classifier language like English where pluralization is obligatory.

Nepali has innovated more than 200 numeral classifiers presumably due to geographical contact with Tibeto-Burman and Austoasiatic languages although it is an Indo-European language like English. It may be the reason why Nepali is reluctant to use plural morpheme for the agreeing targets of INANIMATE COUNT nouns in appropriate contexts (cf. 32 and 33). Or else, it may be that both the processes of obligatory pluralization of Indo-European and the obligatory unitization of numeral classifier languages may be in conflict to assign both the plural and singular morphemes in the targets of this class of nouns.

 

4.3.2 Prototype extension of plural as honorific

Plural morpheme for singular referent has been the original strategy of honorific in the Indo-European language Nepali. This strategy of honorific is also shared by other Indo-Aryan sister languages. We have already said that number is a grammaticalization of QUANTITY. MORE IS UP is a common embodied experience. This may be the reason why typically the plural morpheme was chosen for the grammaticalization of honorific structure in Indo-European. Nepali first applied this strategy in the default grade (LGH) of the human class. Here, both the feminine and the nonfeminine plural forms were chosen to derive the first honorific (MGH) grade. Now the historical feminine plural is lost from the standard dialect although it is still in vogue in the dialects of the Karnali river basin, the form is retained in the honorific singular of the first honorific class (MGH). This characteristic Indo-European strategy of honorific grade became lower when other two special grades of honorific were innovated by the language.

In the historical period when many petty kingdoms were in existence, the royal courts needed another honorific class and so the second honorific (HGH) was innovated in the first half of the 18^th century (Chalise 2005). The feminine plural in the verb inflection is not used in the grammaticalization of verb structure of the second honorific, but Singular-for-plural rule of honorific was retained in the nonfeminine forms of NP targets. When the third honorific (royal) was innovated probably during the 19^th century with a special verb structure, feminine marking in the typical feminine targets was also replaced by the extended use of nonfeminine plural morpheme resulting in the loss of feminine gender in the royal class nouns.

Plural morpheme is implicationally associated with power. This explains why plural morphemes are not used to refer to the royal family members recently after the end of kingship.

4.4 Grammaticalization of Special honorific grades

The verb forms of special honorific grades are derived from biclausal nominalized structures. The second honorific grade (HGH) is derived from a nominalized underlying embedded clause and the verb of the matrix clause by bracket pruning, but the verb structure for the 3^rd (royal) honorific grade is derived from a biclausal coordinating converbial construction through bracket pruning and reanalysis.

 

4.4.1 Second honorific (HGH)

The second honorific structure (HGH) is characterized by a special compound verb structure of the following schema:

 

(42) V-<nu>­ + <ho/hu->/ <b̤ʌ-> + INFL.3SG.NF.

(43) ama          au-nu               ho-l-a

[mother     come-INF]      [be-PROB-NF.SG]

‘Mother may come’ [literally, ‘the act of mother’s coming may happen’]

(44)  ama         au-nu               b̤ʌ-y-o

[mother    come-INF]      [become-P-NF.SG]

‘Mother came.’  [literally ‘the act of mother’s coming happened’]

(45) ama          au-nu               hu-n-tshʌ

[mother     come-INF]      [become-HAB-NF.SG]

‘Mother comes.’  [Literally ‘the act of mother’s coming happens’]

 

The literal meaning of the structure is: ‘the act of something is/was/ may be done’/ ‘some act happens’. This structure is evolved clearly through an image schema extension and reanalysis by ‘bracket pruning’ between the two S-bars, because if the brackets are imagined to have existed, the corresponding literal reading of the resulting sentences exists in the language. The comparison, analogy and leveling between the sentences of this structure and other clauses typical for the default grade and the first honorific grade (MGH) established the so-called ‘subject agreement’ between the subject of the first clause and the newly emerged bracket pruned second verb. This gives the explanation why there is total absence of Plural-for-singular strategy in this honorific grade.

 

4.4.2 Third (royal) honorific

The third (or royal) honorific is characterized by the following structures:

 

A. Special light verb

The light verbs typically used in this grade are derived from the reanalysis of nominalized possessive construction. The possessive postposition appears after the semantic subject with transitive verbs and it is found deleted with intransitive verbs. The light verb makes the sentence superficially transitive.

 

(46) shahdzyada (ko)               sʌbari hu-n-tshʌ

Prince          (POS)           travel   BE-HAB-IS

‘The prince comes/ goes’ [Literally, ‘the prince’s visit happens’]

 

B. Special compound verb

The schema in (47) shows that in this structure the ‘vector’ (Hook 1974) or the grammaticalized verb of the compound verb always comes from the Persian root <bʌks> ‘gift’. This structure is derived from the reanalysis through the bracket pruning of an underlying coordinating converbial construction. The converbial interpretation also coexists in the language.

 

(47) V-<i> + <bʌks-i>/<bʌks-ʌ> + INFL

(48a) shahdzyada        baʈʌ     lekʰ-i                bʌks-i-yo

Prince                 ABL    write-CONV   <bʌks>-DV-3SG.NF.P

‘The prince wrote’ [literally, ‘The prince gave a gift or did a favor by writing’],

(48b)  shahdzyada       baʈʌ     lekʰ-i                bʌks-e-nʌ

prince                ABL    write-CONV <bʌks>-P-3SG.NF.NEG

‘The prince did not write’ [literally, did not give a gift or favor by writing’

 

4.5 Interaction with case

A third person subject of the transitive verb is usually followed by the ergative marker <le> in the default and other two honorific classes, but in the third honorific the ergative postposition <le> ­is replaced by the ablative postposition <baʈʌ>, the dative postposition <lai> changes into the locative <ma>; and the comitative postpositions <sitʌ> and <sʌ̃gʌ> are replaced by <ka saath maa>.

When the light verb allows transitive option, a royal subject may be followed by the ablative one. Dative case is limited to the animate referents in general.

 

4.6 Extension of classifier for gender distinction

It is interesting in Nepali that even the general classifier for count nouns <oʈ-o> inflects for number and gender. However, through analogy the masculine singular form <oʈ-o> is not commonly used in the standard dialect. In most of the dialects the feminine morpheme <i> of the classifier is retained in plural although normally the targets for the feminine change into nonfeminine plural.

 

4.7 Pronominal gender (+classifiers)

An interesting thing in this ‘pronominal gender’ system is that even the human baby is mapped by the nonhuman anaphora until the sex is identified (see 30).

 

4.8 Gender deviation for affection

There is often a tendency to change gender from feminine to nonfeminine and vice versa out of affection. As already noted above, parents can refer to their daughter by using (31a) or (35a) and their son by using (31a) and (35b) when they wish to express love on their faces.

 

4.9  Grammaticalization of demonstrative

There are two sets of 3^rd person pronouns in Nepali. One set is reserved clearly for human while the other set which is developed from the grammaticalization of demonstrative pronoun, does not distinguish humanness; it is rather inclusive (cf. 28 and 29).

5. sociolinguistics

5.1 Gender loss in social dialects

Nepali belongs to the northwestern branch of Indo-Aryan (Turner 1931). It is historically and linguistically related to the Pahadi group of the New Indo-Aryan sub branch of the Indo-European family. It shares most of its characteristic lexical and grammatical features with the western and northwestern groups of New Indo-Aryan languages, but due to its present location where it is in contact with the Bodish branch of the Tibeto-Burman languages on the north, eastern group of New Indo-Aryan languages on the southwest and with other Himalayish group of Tibeto-Burman languages on the east, it has innovated many features. Among other features the one which is of most relevance in this paper is gender loss in the sociolects of Nepali spoken by bilingual Tibeto-Burman speakers and their monolingual children. The process of gender loss is more prominent in the speech of Darjeeling dialect where the majority of the speakers have come from the Tibeto-Burman family background. Priestly’s source of Nepali data is likely to have come from one of these groups. The neighboring eastern group of Indo-Aryan languages like Bangla, Oriya, Rajbamshi and Maithili has already lost gender distinction; therefore gender loss in social dialects is more prominent in the east than in the west.

Gender loss became a characteristic of the royal register probably because of the predominant roles played by ladies most of whom came from different Tibeto-Burman mother tongue backgrounds in the royal court.

 

5.2 Gender shift

Grammatical gender is found in the inscriptional Nepali as well as in the dialects of the far western and mid western regions of the country. However, standard Nepali has undergone a shift in gender from the grammatical to semantic human vs. nonhuman.

 

5.3 Emergence of honorific gender

Nepali has developed honorific as a new set of subgenders. This may be the result of more than fifty petty kingdoms within the present territory after the 14^th century.

 

6. conclusions and issues

Following conclusions and issues come out from out from this paper:

1. Nepali is not a genderless language as reported by Priestly (1983) and taken by Corbett (1991) without verification. It has not only retained sex-genders, but also developed honorific-genders in the human class. There are eleven noun classes supported by eleven different agreement patterns (see figure).
2. Gender assignment in Nepali is ‘strictly semantic’.
3. Lucy (1992) puts a gender prominent Indo-European language like English as typologically distinct where pluralization is characteristic against those numeral classifier languages like Yucatec where unitization is characteristic. However, although Nepali is an Indo-European language to have characteristic pluralization, it has innovated numeral classifiers where Lucy would expect unitization. It is interesting to note that Nepali is often reluctant to use plural targets to agree with inanimate plural nouns. There is an example where even the nonhuman animates in the plural environment are supplied singular targets to agree with.
4. Another interesting feature for Lucy (1992) may be the general classifier for COUNT nouns in Nepali which inflects for both gender and number. The use of classifiers for gender distinction is a unique feature not commonly found in languages.
5. Nepali has based nominal agreement on human vs. nonhuman classes against animate vs. inanimate classes of Indo-European and Indo-Aryan sister languages. Feminine and honorific are the subgenders of the human class.
6. Corbett’s (1991:226) Agreement Hierarchy should also include interrogative pronouns through Nepali data.
7. Feminine and nonfeminine contrast is not a characteristic of Nepali pronouns.
8. If gender assignment is established by the choice of anaphoric pronouns and not only the bound morphemes of the target as in the case of English gender assignment proposed by Hockett (1958), then even the agreement between the nouns and their choice of numeral classifiers may be categorized as gender (Craig Ed. 1986 and Dixon 1986) although Corbett does not like to bring classifiers in gender. However, Grinevald (2000) tries to see classifiers and gender marking morphemes as two extremes of a continuum where numeral classifiers make one extreme and the gender marking bound morphemes make another extreme. As an Indo-European language, Nepali began its nominal classification with bound gender morphemes and as a neighbor of classifier languages it has innovated features of another extreme. However, Senft (2007) has his own reservations.
9. Agreement class in Nepali is a function of regularity and constraints on feminine and plural markings in agreement targets.
10. Loss of gender distinction in pronouns in Nepali (see 26 and 27) is against the expectation of Priestly (1983) but follows Corbett’s (1991:226) Agreement Hierarchy.

LIST OF ABBREVIATIONS

ABL = Ablative

CL = Classifier

CONV = Converb

DV = Default vowel

EMP = Emphatic

ERG = Ergative

F = Feminine

GEN = Genitive

HAB = Habitual

HGH = High grade        honorific

IMP = Imperative

INF = Infinitive

LGH = Low grade honorific

MGH = Middle grade honorific

N = Noun

NEG = Negative

NF = Nonfeminine

NHUM = Nonhuman

NP = Nonpast, Noun phrase

OPT = Optative

P = Past

PL = Plural

POS = Possessive

PROB = Probabilitative

REL = Relative

SG = Singular

V = Verb

VP = Verb phrase