Table 8 3.2. Sentences with non-third person object
Related Figures (87)
Table 1 Purum person-marking suffixes Table 3. Regressive vowel harmony Table 2 Purum vowels vowel /e/ in the prefix if the stem contains /e/. This is nicely shown in examples (1) and (2). The behaviour of the back vowels is the same as that of the front vowels. Table 3 below presents a pattern of regressive vowel harmony and this applies both to the subject and object prefixes, as we can see from (18), where we find kini-; here, the combination of the Ist person singular subject prefix and the object prefix are realised with the vowel /i/ in harmony with the vowel of the stem minrey ‘love’. 2.3. First person singular prefix examples (3c)-(3d). The elicited data we use here are taken from G Singh (2003). The “inclusive” and “exclusive” distinction of first person plural pronouns and their corresponding prefixes refers to first person dual or plural pronouns. A pronoun is said to be “inclusive” where the addressee as well as the speaker is included while an “exclusive” pronoun does not include the person being addressed (Crystal 1999: 158). This distinction is seen to be applicable for Purum. When the first person plural subject pronoun includes the listener(s), the verb takes ni- “1PL.INcL’ when the listener(s) are not considered as a part of the subject argument, the verb takes kin- ‘lpL.excL’. The same prefixes indicate possessor pronouns when occurring on nouns. This plural inclusive/exclusive distinction is still maintained when a noun is prefixed with ni- or kin-. Examples (5a)—(Sb) are with kin terms and (6a)—(6d) are with verbs. Example (6c) turns out to be ungrammatical because the first person plural inclusive subject co-occurs with a second person in the object position. Note that in (6d), the verb is inflected with kin- ‘1PL.ExcL’ twice 2.6. Second person plural prefix present a detailed analysis of the Purum case system for reasons of its exceeding the scope of this paper. Table 4 Subject and object marking on the verb Table 5 Aimol subject and object person-marking affixes 5. Ditransitive verbs In Purum the verb can also be inflected with an indirect object prefix. In (23) we see a first person subject and second person indirect object both marked. If the indirect object is a third person, its prefix on the verb is represented as zero, as in (24). Table 6 Negative suffixes This section deals with negative sentences in which a negative suffix is added to the verb but there is no person-marking prefix on the verb. Purum uses different negative suffixes that agree with subjects in person and number as shown in Table 6. It is interesting to note that both the first person singular negative suffix and the first person plural exclusive negative suffix end in the velar nasal /n/ and the first person plural inclusive negative suffix uses a different form as in No. 2 of Table 6. The third person negative suffixes end in the velar stop /k/. Note that the negative sentence uses the same object prefixes as found in the assertive sentence; the non-third person object prefix is na- and the third person object prefix is represented as @. Consider the examples in (25). Table 1 The eighteen letters of the traditional Meetei Mayek alphabet, arranaed in the traditional bodv-part order Table 2. The existing twenty-seven letters of Meetei Mayek Table 4 Meitei Mayek final vowel markers Table 3 Syllable-final consonants, the lonsum iyek In syllable-final position, seven consonants and one vowel have a contracted form which is called lonsum iyek. These are called “unreleased” in most descriptions of this script. They are given in Table 3. Table 6 Meitei Mayek numerals Table 5. All other punctuation uses standard international punctuatior marks. Table 5 Meitei Mayek punctuation Three punctuation marks, including one tone mark, are included in Unified Meetei Script. These are called khudam iyek; they are listed in are ignored completely on this diagram, all invisibly clustered together on the short stem that is labelled “Tangkhul’”. Any linguistic reader would assume that all varieties of Tangkhul are mutually intelligible. Figure 2 Classification by the “Linguist’s” and the “Tribal” definition Three Meanings of “Language” and “Dialect” in North East India 41 Table 1 Languages officially recognized by states as “languages”. In India, people who speak a “dialect” may even be in danger of losing the legal protection that the Constitution gives to “languages”. Most linguists would like to encourage all forms of speech, even those that are not written, to be treated as honourable ways to talk, but we need to be clear about what we and others mean by the words “language” and “dialect”. When we are talking or writing for other linguists, we should be able to use the linguist’s definition with confidence. If we are listening to advocates with a political agenda, we should never forget that they are almost certainly using the words in some other way than we usually do. If we want to support the linguistic minorities of North East India, we should, whenever we have the chance, insist that the way people talk is important, and that each person’s way of talking deserves respect, whether or not it is written. In principle, whether something is called a “language” or a “dialect” should not matter. But it matters a great deal if, by having their language dismissed as a “dialect,” people are deprived of their rights to use their own form of speech. Table 1 Pronoun forms across Boro-Garo Table 2 Modern Boro-Garo demonstratives The demonstrative pronouns in Table 2 evidence a similar disparity in forms across the modern languages as both u and bV forms are found. One possible etymological source for the BG demonstratives could be found in the Tibetan demonstrative system. Franke writes: Table 3 The PTB transitive verbal suffixal paradigm (from DeLancey 1988) Table 4 Indo-Aryan demonstratives Table 5 Cognate classifiers across BG and reconstructed PBG forms While Joseph and Burling (2006) provide reconstructed forms for these elements, they offer no explanation of how the system has changed since PBG. For example, modern Garo has replaced the PBG three- dimensional object classifier *tai with a more recently innovated form roy-, which is derived from the noun roy ‘stone’. Other languages have simplified the PBG classifier system such that Boro has lost the residual category, Rabha the 3p object category, Deuri the animal and 1p object categories while Dimasa has lost the 1p object and residual categories (Wood 2008). 2.4.1. Nominative case Table 6 Case-ending suffixes across Boro-Garo Table 1 Jinghpaw prefinal SFW agreement endings The glottal finals in the 1“ singular forms reflect earlier *k (PTB final *k > 2 is aregular Jinghpaw sound change), thus the voiceless stop forms are the original nasal suffixes subject to some kind of devoicing process. Since this process has no synchronic phonological motivation in either language, it must be of some reasonable time depth. The voiceless forms are followed by the simple final SFW ai, but the nasal forms are followed by the inflected form with person agreement (4)-(8). 2.2 Hierarchical agreement and External Possession Note 2™ person agreement in (19), indexing the possessor of the 3"¢ person subject; 1*t person agreement in (20) is with the possessor of the object. Table 2: Nocte intransitive agreement paradigm Or they may be combined with one of a set of morphemes which. like the initial elements in Jinghpaw SFW’s, consist of a single consonant. and indicate various tense/aspectual and other verbal categories (21)- (25). Hierarchical agreement occurs elsewhere in Tibeto-Burman (Delancey 1980; 1981a; 1988; 1989), but it is nevertheless noteworthy to find it in these two systems which show so many other correspondences. A striking feature of the Nocte paradigm is explicit marking of the grammatical category called inverse. The h- morpheme indicates that agreement, rather than being with the subject, as is typical in Indo-European languages, is Another striking trait which Nocte shares with J inghpaw is hierarchical agreement: the verb agrees with a 1‘ or 2" person in preference to a 3", regardless of which is subject or object (26)-(29). 4. A comparison of the Nocte and Jinghpaw SFW systems Table 3 Nocte transitive agreement paradigm Table 4 Jinghpaw and Nocte cislocative paradigms The Jinghpaw plural forms are distinguished not by distinct agreement suffixes, as in Nocte, but with the plural md- prefixed to the SFW, as in ex. (30). One form clearly is cognate, the r- morpheme which in Nocte indicates cislocative, that is, motion toward the deictic center. It is “only possible in present tense” according to Weidert, and in Jinghpaw has tense functions in the realm of future/progressive, but yet also has a clear cislocative force (DeLancey, Diehl et al. 1978). The formal correspondence is evident in Table 4, with Jinghpaw sa ‘go’ + r- and Nocte ka(t) ‘go’ + r- both meaning ‘come’. So we have correspondence in form, meaning, and combinatorial behaviour between not only Jinghpaw and Nocte r-, but the whole SFW construction in which they participate. Table 5 Correspondence of pronominal and agreement forms with some sort of syllabic prefix (Bauman 1975), but these are secondary developments, and the presence in both languages of the unchanged original pronominal forms shows that nothing of the sort has occurred in the history of this branch. Table 6 Source of Jingoaw SFW forms Abbreviations Table 1 Tangsa tones Table 2 Comparison of Tangsa words have the form ingat for ‘one’. Finally the three sources for Moklum are listed. One feature of Moklum is final -/ on ‘fire’, ‘sun’ and ‘good’. This feature is partially shared by Havi, Mosang and perhaps Morang. Figure 1 Location of Kharang Kong village (approx 27°22’ N, 95°48’ E) Table 3 Correspondence of Cholim simple vowels with Lochhang Table 4 Cholim agreement markers Figure 2 Wave form for me tadyoeq gep nyen ang Table 5 Cholim imperative/hortative markers Table 6 Lochhang agreement markers As can be seen in Table 6, Lochhang agreement markers are treated as clitics. They are prosodically closely bound to the verb, as shown in (21), and in Figure 3, the wave form for (21). When comparing Figure 3 with Figure 2 above, we can see that there is no gap between the future clitic ma@= and the verb, nor between the verb and the agreement particle. Both figures occupy nearly the same time length, and the difference in prosody can clearly be seen. In all of our recordings of Lochhang made so far, we have observed the same kind of prosodic structure of agreement markers that is observed here, and in general one of the more salient differences between Lochhang and Cholim is the prosodic status of the agreement markers. Table 8 Ditransitives in Moklum (after Das Gupta) The ending -phu in (28) suggests that final -v marks the 2™ singular and that in Moklum the 2™ person singular past marker might be tu, missing from Table 7. In (30), the 2" person agent is marked by -a, called nominative by Das Gupta but here glossed as ‘agentive’. Its presence with the verb ‘beat’ suggests a stronger agentivity/transitivity for ‘beat’ than for ‘call’ in (31). Monotransitives behave in exactly the same way, as we can see in (29)-(31). Table 10 Imperatives with 1St person objects in Moklum (after Das Gupta) Thus, the whole imperative system appears to be as in Table 11. Table 9 Imperatives in Moklum (after Das Gupta) There is also hierarchical marking of imperatives when 2" person acts on 1*' person, as shown in Table 10, with sin ‘look’. Table 11 Imperatives/hortatives in Moklum Conclusion Table 1 Definite and indefinite acs (GruBner 1978: 123-4) Table 2 Constituent Order within Adjectival Attribution vs. RCs The Karbi examples (12) and (13), where the modifying RC precedes the head noun, can be contrasted with (5—8) above, where the modifying derived adjectives follow the head noun. Table 3 Nominalization patterns in adverbial subordination actually a semantically motivated hierarchy of pairings of adverbial subordination type and syntactic construction (i.e., nominalized or non- nominalized)? What is the semanto-pragmatic difference between a nominalized and a non-nominalized construction for the same subordination type? Table 4 Summary the potential pragmatic or sociolinguistic motivations for choosing one construction over another. Figure 1 réluiu (Galo) village, West Siang district, Arunachal Pradesh, with moodii-rikd (jhum fields) in background and terraced isi-rikd (wet fields) in foreground Figure 2 Ziro, Apatani plateau; Lower Subansiri district, Arunachal Pradesh Figure 3 View of Upper Assam from the West Siang district border, with Mising and Bodo-Kachari villages in distance Figure 4 Environmental source of topographical deixis Figure 5 Environmental extension 1: Riverine orientation From the topographical source illustrated in Figure 4, a number of environmental extensions are also observed. The first is riverine orientation: in this extension, upward is transferred to upriver, downward to downriver, and same/unknown level is effectively extended to mean not along a river course (Figure 5). Figure 6 Environmental extensions 2: Compass orientation Topographical Deixis and the Tani Languages of North East India 14 Table 2 Galo directionals encoding topographical deixis The set of directional predicate derivations (“directionals” for short) encoding topographical deixis in Galo are presented in Table 2. The Table 1 Galo verb roots encoding topographical deixis The set of Galo verb roots encoding topographical deixis are presented in Table 1. An example of use is in (2). Note that one of the three verb roots is polysemous with two other deictic domains (motion towards a reference point and motion into an enclosed space); note also that although the English translations are necessarily somewhat elaborate, these are among the most frequent and ordinary verbs used for expressing motion to a goal in Galo. 5. Topographical deixis in other Tani languages Table 4 Galo demonstrative postpositions exhibiting topographical deixis Table 6 Sample set of terms employing topographical deixis in Pagro Mising Table 3 Geminate and non-geminate forms of PD and SD: Root words along with their syllabic representations Among the PD forms shown, when -a is suffixed to a root word, gemination of the medial consonant and deletion of the word-final vowel are observed. But we see that the suffixation of -a to an SD root word does not bring about any change to the form of the word. Notably, the It is evident from the examples given above that the root words oma ‘daughter’ and fab: ‘snake’ in both the dialects are same, but the combination of -a with the root words bring about morphophonemic changes in PD and not in SD. The alternation is further exemplified in Table 3. In words ending in u include the structures VV, CVV, CV.CV, VC.C\ and CVV.CV. Consider the following examples: As above, words which have light-light syllable structure (V.CV and CV.CV) as in (12c) and (12d), allow coalescence and as well as gemination when -a is fused with word-final u. On the other hand, words with heavy-light syllable structure allow coalescence but do not trigger word-medial gemination (see (12e)). Here, the fusing of u and -a results with o as the output vowel and it can be represented as V, (high, back) + V, — V, (mid, back) or u + -a > 0. Nfanneyvllahir wrarde unth VV and OVV otricrtire (199_h) naithear Examples (1la—11g) above exhibit words ending in 7 with various possible syllable types (VV, V.CV, CV.CV, VC.CV, CVC.CV, CVV. CV. and CV.CVV) in PD. Here, we have both monosyllabic and disyllabic words. We notice the presence of light and heavy initial and final syllables in the given disyllabic words. Words with light initial and final syllables (VC.V and CV.CV, as in (11b—11c)) both allow coalescence and produce word-medial gemination when -a is attached. On the other hand, a disyllabic word with heavy initial syllable and final light syllable (VV, CV.CV, CVC.CV, and CVV.CV) allows coalescence but not word- medial gemination. The fusing of i and -a usually gives the resultant vowel e, and this can be represented as V, (high, front) + V, — V, (mid, front), or i + -a — e. It is worth noting here that a monosyllabic word in PD which is in diphthongal form (11a) or a disyllabic word in which the word-final syllable is a diphthong (11g) neither coalesces nor exhibits word-medial gemination; rather, we notice the insertion of the palatal continuant j in between the root-final vowel and the suffix -a. Note: WMG = word-medial gemination, IS = initial syllable Table 4 Morphophonemic changes in PD root words (ending in short vowels) with various syllabic structures - The source of Boro data is Bhattacharya (1977). Table 1 Comparison of Assamese determinatives and Boro classifiers Table 2 Comparison of Assamese determinatives and Boro classifiers Even some such morphemes are common to both Boro and Assamese (Table 2); this would suggest that the two systems may at least partly share a common origin. Table 3 Determinatives in Assamese A short list of the determinatives found in Assamese is given in Table 3, together with illustrative examples of their use. is, however, an important difference between NPs with an interrogative determiner and NPs with a demonstrative determiner. In the latter case, the ptv can be affixed either to the determiner or to the head; in the former case, the ptv can be combined with the determiner only (9)- (11). Table 4 Determiner function of three different classes of pronouns Table 5 Use of ptv with quantifiers 3.6 Genitive forms Compare the homophonous intransitive verb as- ‘exist’ in an existential sentence which exclusively co-occurs with a 3™ person subject. Table 2 Finite inflectional paradigms of primary copulas Table 3 Constituent orders of copular sentences To recapitulate, the primary copular sentence of both the subtypes follows the canonical SO(V) order of the language. By contrast, the quasi copular sentence exhibits SVO order. While the inversion of CS and CC in a primary copular sentence may affect its interpretation or render an ill-formed sentence, it does not affect the basic proposition of a quasi copular sentence. Table 4 Case marking of CS and CC in primary copular sentences However, one of the striking features of a quasi copular sentence is that only a NP, unmarked for case, can occur as its CC as in (10), (13). Table 1 Case markers which immediately follow the noun may be attached to the base indirectly, may be mediated by a Layer I element and/or is invariant for all nouns and the same for both numbers. Layer II elements may be either agglutinative suffixes or analytic particles and sometimes even clitics. Layer III is mediated by a Layer II element; it lacks morphophonemic variants and often has a transparent connection with an independent word; it is semantically more specific than a Layer II element. All of the Hajong case formatives listed in Table 1 above fall into Masica’s category of Layer II affixes since they are invariant for all nouns and the same for both numbers. These case formatives are not phonologically dependent on the noun. They can occur directly after a noun or after its classifier. In this paper, I will usually separate the case formatives from the preceding noun words. The exception is when the case formative follows a pronoun in a form which cannot stand alone. Here, the case formative is hyphenated. Table 2 Case markers which follow the genitive The markers for specific location such as ‘on top of’ or ‘underneath’ can occur with or without the genitive intervening. The genitive marker comes with a change of meaning, as shown in examples (1) and (2). Hajong also has Layer HI elements which specify the function of a noun phrase. The benefactive and comparative markers are mediated by the Layer II genitive marker, as shown in Table 2. Table 3 Cross-linguistic case marker comparisons common in NIA languages to use one marker for both of these functions (Masica 1991: 365). The two forms, ge and gon, are synonymous and in free variation although a given speaker will choose one or the other. Examples (3) and (4) illustrate that, subject to animacy conditions elaborated on below, both ge and gon can be used for the accusative case, i.e. the object of a transitive clause. Table 4 Cross-linguistic comparison of the accusative/dative formative One possible cognate for the Hajong gon is the Sylheti!! classifier for animate nouns which is -gu in the singular and -guin in the plural (Lloyd- Williams, personal communication). The morpheme -in is the most common plural in Sylheti. Although these classifiers in Sylheti are not restricted to the accusative or dative positions, they are used exclusively for animate nouns as the Hajong accusative/dative is used only for animate nouns. The Hajong gon could also be related to the archaic Bangla plural morpheme gon. Table 6 Possible cognates for Hajong case markers Khasi has the same nominalizer noy- which changes a verb to a noun, as shown in example (9). It shows the derivation of nominal ‘beater’ by prefixation of noy- to the verb so? ‘beat’. 244 North East Indian Linguistics Volume 3 Table 2 Nominalization of adjectives with Khasi nominalizer iy- Table 1 Nominalization of verbs with Khasi nominalizer ;iy- por- may also be considered a derivational prefix in Pnar which results in the creation of new word, although it does not change the grammatical category. In other words, it is a class maintaining derivational prefix. From the interpretation of the data it has been found that this prefix is used to denote a sense of time frame or duration as illustrated in the following example (11). In the example (11) por- is prefixed to the word k'inna? ‘child’ to derive the word ‘childhood’. A preliminary analysis may be that this prefix serves to create a more abstract noun from a concrete one, in the sense that ‘childhood’ is more abstract than ‘child’. However, further research is needed to fully support this notion.
