Introduction

In some languages, existential, locative and possessive predications are handled by verbs, verbal auxiliaries or copulas with tailored existential, locative or possessive meanings such as ‘be (at)’, ‘live’, ‘have’ or ‘exist’. In other languages, some or all of these functions may be coded by verbs of posture or orientation such as ‘sit’ ‘stand’ and ‘lie down’, among others (cf. several papers in Newman 2002; also Hellwig 2003). Often, existential, locative or possessive predicate selection in the second type of language depends on, or makes reference to, some inherent properties of the focused referent (i.e., the existee, locatee, or possessed), and/or its orientation in space (i.e., in terms of the speaker's construal of the situation). Aikhenvald (2000: §6.2.3) has described this operation as ‘classificatory’, as it basically reflects the same semantic and pragmatic principles governing selection of numeral or nominal classifiers in languages which have them.

In Tibeto-Burman languages, as discussed in particular by LaPolla (1994), it is common to find two, three, four, or as many as seven ‘existential’ or ‘locative’ verbs, selection of which depends at least in part on some attribute(s) of the referent whose existence, location or possession is under predication, such as animacy/mobility, abstractness or concreteness, or degree of independence or containment, etc.; i.e., they appear to ‘classify’ referents in Aikhenvald's (2000) sense. In the Tani languages 128 North East Indian Linguistics (see Figures 1–2), we find as many as five such verbs; henceforth, I will describe these verbs (for Tani-internal purposes) as special intransitives. My main purpose here will be to provide, to the extent currently possible, an overview of the verbs involved and their functions in different Tani languages.

Introduction

In 1904, Wilhelm Schmidt gave comparative evidence in order to establish the notion that Khasi was a member of the Austroasiatic family. In support of this, he proposed many etymologies linking Khasi to the Palaungic languages, Khasi's nearest neighbours to the East (Schmidt 1904).

Over a century has now passed, and with the better knowledge of Austroasiatic which we have today, we can see that Schmidt's findings were fundamentally correct. We notice that a few of his etymologies require corrections, or in some cases rejection for specific reasons; but, more importantly, we now have numerous new etymologies, inaccessible to him at the time, which support his general claim. For example, a rapid look at a reconstructed vocabulary of the Waic languages (Diffloth 1980) will make the point clear (Waic being a sub-branch of Palaungic).

This being the case, several other questions present themselves. Geographically, Khasi is a linguistic island surrounded today on all sides by Tibeto-Burman and Indo-Aryan languages. Its Palaungic relatives are about five hundred kilometers away as the crow flies. Moreover, the high mountain chains and deep gorges that stand in the way in Northern Burma are oriented for the most part in a North-South direction, making East-West travel between the Palaungic and the Khasian regions even more difficult than the distance indicates. How then did the languages come to be where they now are? Languages evidently have been spreading one way or another, but which ones, by which ways, and in what times?

Half a century after Schmidt, Robert Shafer wrote an article concerning these unanswered questions (Shafer 1952).

Introduction

Atong is a Tibeto-Burman language spoken in the beautiful mountainous region along the Simsang River, between the villages of Nongal Bibra and Baghmara in the Garo Hills in the state of Meghalaya, India. There are an estimated 15 to 20 thousand speakers of Atong. This estimate is given by the Atong people themselves. Officially the Atong are grouped together with the Garos. Garo is the name given by outsiders to the Tibeto-Burman people living in the Garo Hills and adjacent areas who speak a group of dialects also called Garo. The Atong language, which has no official status, is mistakenly considered to be a dialect of Garo by the majority of people living in the Garo Hills. The image that Atong people seem to have of themselves is that they are a distinct group among the Garos. Some Atong do not identify themselves with the Garos at all.

The Atong speaking area can be roughly divided into two dialectal areas, viz. a Nongal-Rongsu area (north) and a Siju-Baghmara area (south). Due to the mobility of people in the area and of men in particular, Atong speakers are aware of the dialectal differences which are mostly lexical and prosodic in nature and show very little grammatical differentiation. The dialects are perfectly mutually intelligible. There is some dialect mixing in the area. The Atong are a matrilineal and matrilocal society. When a boy marries he will move into the house of the family of his wife, even if she lives in another village. Further dialect mixing occurs at the market where people from different villages come together.