Introduction

The aim of this study is to provide a morphological and syntactic description of personal pronouns in Madhav Kandali's Ramayana (MKR). Prior to the main discussion on grammatical intricacies to be followed hereon, a brief background of the MKR is in order to authenticate the relevance of this study.

MKR has been attributed to the 14th century, (Sharma 1996; Neog 1985), a period of great significance in the history of the Assamese language, where the first phase of renaissance in religious, cultural and linguistic field was taking place. Around five court poets, including Madhav Kandali, produced Assamese texts during this period. Madhav Kandali translated the entire Sanskrit Ramayana composed by Valmiki into Assamese at the behest of the king Maha-Manikya of Tripura (Medhi 1936: 70). Though Kandali was an accomplished scholar of the Sanskrit language, he did not use the style of Sanskrit in his translation. He wrote it in an easy and simple, yet sublime style to cater to the needs of spiritual leaders to religious enlightenment of the mostly illiterate masses.

It is believed that over the 6th and 7th century, AD, the spoken form of the Assamese language began to develop and spread among the people. It got its written form during the 12th and 13th century, AD (Saikia Borah 1993). The specimen of this period is Carya, written by the Buddhist Siddhacharyas on palm leaves (Tāl pāt), hailing from the different parts of ancient Kamrupa.

Overview of the Mon-Khmer (MK) languages of India (Meghalaya, Assam) and Bangladesh

The MK group of languages spoken in NE India, especially in Meghalaya since the formation of this State in 1972, is currently known as Khasi or Khasian. These two terms correspond to different facts and to an unsettled question. Standard Khasi has become the lingua franca and the written language of the MK eastern area of Meghalaya since British colonisation. This situation reflects the political and socio-economic leadership of the Khasi group over other MK groups in Meghalaya and also the isolated cultural situation of the MK group inside NE India. As a state of India, Meghalaya has two parliamentary constituencies: a Garo one and a Khasi one. Any War, Pnar or Lyngngam person would say he is Khasi as some kind of “national” identity, different from any other neighbouring Tibeto- Burman (TB) or Indo-Aryan (IA) identity. On the other hand, “Khasian” is not an empirically defined term from the viewpoint of MK linguistic classification. The chapter in Grierson (1904) on “Khassi” and its Synteng (that is Sutnga Pnar), Lyngngam and War “dialects” provides basic lexical lists of 200 items and two translated Bible texts in those so-called dialects. This material shows important lexical dissimilarities and deep morphosyntactical differences. Those languages are labeled “corrupted dialects” of Khasi by Roberts (1891), who provided the data. Those groups and what remains of the former dialects of Khasi: Mylliem and Khyrim (or Khynriam), have never been thoroughly surveyed.

Background

The membership of the Boro-Garo branch of Tibeto-Burman has been clear, at least since the 1903 publication of Vol. III, No. II of George Grierson's Linguistic Survey of India. This volume was devoted to the “Bodo, Nāgā, & Kachin” languages and it included a map, “Bodo Group,” that is still the best map of these languages that I know. This map is reproduced here as Figure 1, but with the modern names for the languages added. Even the subgrouping within the Boro-Garo branch has received a wide consensus, although this has more often been due to copying previous guesses, than any serious new inspection of the data (Burling 1983, Egerod 1980, Hale 1982, Shafer 1966–74). Enough new data on these languages has now become available to invite a fresh consideration of their relationship.

The languages that are presumed to belong to the Boro-Garo group are listed across the bottom of Figure 2, and the forks in this diagram summarize my best present judgment about how the various languages are related to one other. The forks are numbered roughly from the top to the bottom of the diagram, and they represent splits between smaller groups within Boro-Garo. After describing some features that are characteristics of most or all of the Boro-Garo languages, I will point to the evidence that supports each of these splits.

Introduction

In this paper we will examine poetic forms of four languages spoken in Upper Assam and Arunachal Pradesh, India: Nocte, Singpho, Tai and Tangsa. As far as we know there are no words that equate to English ‘poem’ or ‘poetry’ in any of these languages, though each of the languages has a word for ‘song’, and has words for different song styles, such as dance song or antiphonal (call and response) song. Except for the Ahom texts in §5 (which are taken from manuscripts), the examples in this paper were all performed with melody and sometimes rhythmic accompaniment such as the beat of the pestle on the mortar during rice-pounding (see below §3). All could thus also be termed ‘songs’ and it is not easy to make a distinction in these communities between ‘poem’ and ‘song’. Some styles resemble sprechstimme, a style of song more approaching spoken form, and some styles are more melodic (see List 1963: 9 for a discussion of such different styles). In this paper we will not deal with the musical melody. Poetic devices are also found in story telling, but we will not deal with those here.

Kiparsky (1973: 231) claimed that the “fundamental stylistics of poetry … have existed from the beginning”, based on comparison of poetics in ancient and modern times. We expect that the features of the poetry discussed in this paper represent very old and deep cultural artefacts, and will illuminate our knowledge of these languages.

The Dimasas and their language

The Dimasas are known as ‘sons of the great river’ i.e., the word ‘di’ means ‘water’, ‘ma’ means ‘big’ and ‘sa’ means ‘sons’. The language of the Dimasas is known as ‘Grau Dima’ or ‘Magrau’ (mother tongue). This tribe migrated from the Himalayas and came to the Brahmaputra river valley via the Gangetic Plains and made it their land (Thaosen 2007: 4). The Dimasas are mainly concentrated in some parts of Nagaon, Cachar and Karbi Anglong, the entire North Cachar Hills and in and around Dimapur in Nagaland. According to 2001 Census, the Dimasa speakers are approximately 65,000 in N.C. Hills and 40,000 in Karbi Anglong in Assam.

The Dimasas belong to the Bodo-Garo group within the Bodo- Jingpho-Konyak branch of Tibeto-Burman (Lewis 2009). The Dimasas are multilingual. The Dimasas of North Cachar Hills speak Haflong Hindi which is the lingua franca for the various ethnic tribes of the region. Those from Cachar plains use Bengali which is the majority language there, and those of Karbi Anglong and Nagaon use Assamese for communication. Dimasas in Nagaland speak in Nagamese with speakers of other languages.

The dialects of Dimasa are Hasao, Hawar, Dembra, Dijuwa, Humri, Semsa and Walgong. Hasao, spoken in North Cachar Hills district of Assam, has been adopted as the standard dialect by Dimasa Literary Society in its meeting held on the 9 March, 2004 at Haflong.

Introduction

Most Tibeto-Burman languages have clause-chaining discourse structure (DeLancey 1991), where several clauses are chained together in a sequence with only one finite verb. Lhasa Tibetan (DeLancey 1991) and Lahu (Matisoff 2003) are examples of Tibeto-Burman languages with clause-chaining discourse structure. It has been well attested that clause-chaining discourse structure facilitates serial verb constructions, which, in turn, facilitates the development of grammaticalized verbs in such languages. In this process, grammaticalized verbs may even further develop into prefixes or suffixes. (DeLancey 1991). These grammaticalized verbs mostly develop from motion verbs like ‘go’ and ‘come,’ and postural verbs like ‘stand’ and ‘sit’ (Foley and Olson 1985).

Boro, a Tibeto-Burman language, has both a clause-chaining discourse structure and a serial verb construction. We will talk about these two constructions in detail below. The focus of this paper is a small set of verbs which can occur in the serial verb construction. Some of these verbs have developed a fair amount of grammatical meaning, while others still retain their lexical meaning. Some of them have undergone vowel change too. However, they are still identifiable with their non-serialized main verb counterparts. We will call these verbs ‘serialized verbs’ rather than grammaticalized verbs for obvious reasons. The paper has the following structure: §2 gives a brief background on the Boro language in general, and the dialect on which this work is based in particular.

Introduction

The basis for the current research is our linguistic work among the Hajong community in 2004–2006. The language is spoken by approximately 80 to 100 thousand people (Hajong 2002: 9) in the North-East Indian states of Meghalaya and Assam.

Hajong is classified as an Indo-Aryan (IA) language. It has some degree of similarity with Assamese and Bengali, the two IA languages spoken in the region. At the same time, certain grammatical similarities such as case marking can be found between Hajong and some Tibeto-Burman (TB) languages spoken in the same geographic location (Phillips 2008).

The paper presents the distribution of the vowel and consonant segments, with contrastive pairs, syllable and stress patterns where applicable. At the end of each section we will compare the phonological features of Hajong with the corresponding ones in TB, represented by Garo and IA, represented by Assamese. The scope of the current study covers segments distribution patterns at the syllable level, with some comments on their position within the word.

Analysis of consonants

In this chapter, we will talk about consonant segments of Hajong. A segment is defined as “any discrete unit or phone, produced by the vocal apparatus, or a representation of such a unit” (Loos 1996: S). The next linguistic level above the segment is the syllable. Since our analysis is based mostly on syllable structures, it is important to establish the definition of this notion. A syllable typically consists of a central peak of sonority (usually a vowel) and optional less sonorant segments before and/or after the peak (Loos 1996: S), commonly called onset and coda.

Introduction

Kudeng War is a dialect of Nongtalang-Nongbareh War, one of the three sub-groups of War, itself a nearly unknown, unwritten conservative Mon- Khmer (MK) language. War is spoken by approximately 65,000 speakers mostly in Meghalaya, North-eastern India. Wars are mostly betel nuts planters on slopes and small isolated groups remain in Bangladesh, in Assam and between Mizoram and Tripura states on the former space of an important pre-colonial Pnar kingdom. War belongs to a group comprising Pnar, Khasi, Lyngngam (PWKL) and many “composite varieties”. An overview with two linguistic maps of PWKL and War and also a glossed sample text in War can be found in Daladier (this volume).

War serial verb constructions (SVCs) share common features with SE Asian SVCs, like adversative constructions and serial lexicalization as a complementation strategy. They also exhibit features not yet described which might be shared by other Austroasiatic languages. War SVCs include values expressed differently in a language like English, in active and in passive verbal auxiliations and also in “light verb” constructions; they also include values expressed in English clausal dependencies without realis/ irrealis, finite/non finite or past participial/gerund morphology.

In addition to these values, grammaticalized in their own way, grammaticalized SVCs in War express productively values not grammaticalized in English. They express especially a gradience of active and passive values associated with subjectivity values referring to the speaker (or his subject).

Introduction

Copular constructions generally refer to clause structures where the subject is linked with nonverbal predicates like nominal, adjectival and locative. The purpose of such construction is to express notions like equation, attribution, location, existence and possession (Dryer 1985, Payne 1997). Traditional grammar defines the element that links the subject and the predicate in copular constructions as copula. Following Payne (1997: 111–121), the function of copula can be mainly divided into three: equative, attributive and locative.

Copula can be defined as a term used in grammatical description to refer to a linking verb ….whose main function is to relate other elements of the clause structure especially subject and complement. (Crystal 1980:93)

Masica (1991: 337) states: in Indo Aryan languages, “the function of copula is to identify, define and locate” the Subject NP. Copula in Indo Aryan languages can be overt and covert. It occurs in the final position in the clause structure like other lexical verbs. According to Masica (1991), verbs of this type perform two functions that of copular and existential. Normally copulas can be deleted whereas existential verbs cannot be deleted. Masica (1991: 337-338) further observes that existential verb can be covert in Assamese and Bangla.

Sadri, which is claimed to be an Indo Aryan language (Grierson 1903), also has different copular elements with similar functions, that of equative, attributive and locative. The copula in the present tense can be either obvert or covert. The construction where the copular element remains covert is considered as no copular construction.

Introduction

Usoi is a southern variety of Tripura/Kokborok, a member of the Boro-Garo branch of the Tibeto-Burman language family. It is spoken in the Chittagong Hill Tracts (henceforth CHT), Bangladesh, and in the neighboring Indian states of Tripura and Mizoram as well as in the Rakhine state of Burma. The population of Usoi speaking people in CHT is estimated to be ca. 20,000 (Kim et al. 2007: 9). In India, Usoi is a scheduled tribe and the corresponding population is 1,295 (1981 Census). The population figure is not available for Burma.

Usoi is divided into two major dialects that correspond to two major rivers in CHT: Karnaphuli and Sangu. Although there do exist differences between these two dialects, they are not so divergent that they are mutually unintelligible. My consultant of the present study is originally from Roanchari region in the Karnaphuli basin and now lives in Bandarban, the district headquarters in the Sangu basin. Usoi is highly influenced by Marma, a dialect of Arakanese/Burmese and a lingua franca in the Bandarban district of CHT. As a result, there are many Marma loanwords in Usoi.

Usoi is most similar to Reang/Riang, another southern variety of Tripura/Kokborok; however, there is virtually no linguistic data on Reang/Riang except for Riyang (2007), and the exact linguistic relationship between Usoi and Reang/Riang is unclear. While there are several linguistic descriptions of Kokborok (Karapurkar 1972; Pai 1976; Jacquesson 2003; Jacquesson 2008, etc.) including reliable dictionaries (Debbarma 2001; Debbarma 2002), there are hardly any linguistic reports on Usoi, except for Walsh (n.d.; the 1960s), Kim et al. (2007) and Huziwara (2008).