The Language of the Modhupur Mandi (Garo), Volume 2

Introduction to Volume II

When I first went to live among the Mandis in Bangladesh I had no intention of making a study of the lexicon, but I wanted to gain some fluency in the language and I began to collect words for my own use. As the words accumulated I found it helpful to sort them into rough semantic categories: kinship terms here, body parts there, trees in some other place. Gradually, I became fascinated by, even addicted to, the search for words, and as my lists grew, I kept looking for ways to sort them. I even hoped that a plausible sorting might reveal something about the workings of the language, and not merely reflect my own preconceptions of how words ought to be organized.

My methodology was simplicity itself. When a category grew big enough to become cumbersome, I looked for a way to divide it into smaller but still coherent categories. This worked well enough for some areas. Kinship terms fell naturally into those for men and for women, and for older and younger generations. Mandi names for animals could be sorted, as easily as English words, among birds, fish, land animals and bugs. Vast areas of the vocabulary fell into no such clear categories, however, and there were many hundreds of words that would fit easily in three or four different places. A disturbing number seem to fit nowhere at all. What does one do with a word that can mean 'messy' of possessions or 'confused' of an argument? To extend a word's meaning from messy objects to messy language is reasonable enough, but it makes categorization itself more than a little messy. My only recourse has been to include such words in more than one category, occasionally in three or four. Still, I did try to keep the multiplication of entries under some sort of control. Any native speaker could easily double, and then redouble, the length of most of the categories. Some categories are certainly more complete than others. I doubt if I have missed many Page  vikinship terms except for the odd idiosyncratic term from a dialect that I have never encountered, and I have not included all the Bengali kin terms that Mandis sometimes use as alternates for their own. I certainly have a much higher proportion of mammals than of insects or fish, but it is in the more abstract areas of the vocabulary, verbs and adverbs in particular, where the limitations are most severe. Thousands of words are still out there waiting to be written down.

This volume is made up of several lists.

1. The first and largest part of Volume II consists of my word collection, organized by the categories that emerged as it grew. Most of its words are those used by the Mandis who live in the Modhupur area of Bangladesh, but they share most of their words with other dialects. In order to suggest how the dialects differ, I have given some A˙A chik equivalents, but I have not done this systematically, and anyone who wants to know more about the more "standard" A˙A chik dialect should turn to one of the Garo dictionaries.

The remaining lists are intended primarily for people who want to learn to speak Mandi or another dialect of Garo:

2. A survival vocabulary of about 250 of the most common and useful words. This is intended as a guide for people who want to get started with learning the language. Where the dialects differ, I have done my best to give A˙A chik equivalents along with the Mandi forms.

3. A somewhat longer, but still basic vocabulary list of almost 1000 words. This is meant to give some guidance to people who want to move beyond the most basic level. Different learners will have different needs, of course, and nobody should be confined by my judgement about which words are most important. The words each person needs can always be sought in the full list.

4. The final list of this volume is an English index to the full Mandi word list. Superficially, this may look like an English to Garo word list, or even a short English to Garo dictionary, but it cannot be used without referring to the full Mandi list. The entries on the fourth list will lead you to the Mandi words that overlap in meaning with the English word, but many will not be exact equivalents. Since a single English word may have meanings that are divided among several Mandi words, you will have to look up the Mandi words in the full list in order to find the best word for your needs.

One important way in which the vocabulary of Mandi differs from that of dialects spoken in the Garo Hills is that it has been even more heavily influenced by Bengali. Mandis feel absolutely free to incorporate any Bengali word into their speech that they think their listeners will understand. This means, in principle, that the entire vocabulary of Bengali is available for use by the Mandis. To fold an entire Bengali dictionary into a Mandi Page  viiword list would be absurd, of course, but to include no Bengali words at all would give a terribly false picture of the language as it is actually spoken. The impossible problem is to decide which Bengali words to include and which to leave out.

In an attempt to keep sheer bulk under control, I have taken a somewhat restrictive approach. I have, for example, included only the most common of the Bengali numbers. Bengali has an extraordinarily complex number system and there seems little point in copying that whole system into a work that focuses on Mandi. On the other hand, I have included the Bengali names for the days of the week, which are the only words that Mandis ever use for them. I also include the names of the twelve Bengali months. Mandis now use the western calendar for some purposes, but they continue to use the Bengali months to describe their agricultural cycle, and it would be artificial to leave them out. I welcome words that are used for the daily activities of the farming villages where I have lived, but I omit large swathes of Bengali vocabulary that are used for religion, government and politics, education and learned subjects, and all aspects of urban life. I have included the Bengali words that seem to me to be best established and least avoidable for those who still live in predominantly agricultural villages. I have simply had to use my judgment.

I have spelled borrowed Bengali words to reflect the pronunciation that I have heard from Mandis. This is often quite different from their pronunciation by native speakers of Bengali, and it is even further from a transliteration of Bengali spelling. Some borrowed words are well established and have gained a consistent conventional pronunciation when used in Mandi sentences. Other words vary chaotically. I have included some of the alternative pronunciations, but by no means all. Some borrowings are revealed by their telltale phonology, but many are not. My own skills in Bengali do not go beyond a sort of rudimentary pidgin, so I have generally had to rely on my consultants to know which words have been borrowed. Their reports, unfortunately, are not always consistent. Occasionally I have been given synonymous pairs, only to have one Mandi speaker attribute one of them to Bengali, while another speaker points to the other. More often, fortunately, Mandis know perfectly well which words are borrowed. They have had plenty of experience speaking to Bengalis so they know which words work and which do not. It is more likely that I have failed to recognize some words as having come originally from Bengali than that I have mistakenly attributed unborrowed words to Bengali, though it is all too likely that I have made both kinds of errors.

As with Bengali, I spell words that are taken from English in the way that I have heard them pronounced by Mandis, and as I pronounce them myself when I speak Mandi. As a speaker of English, I have much more Page  viiiconfidence in my own ability to spot borrowings from that language. "<B" indicates that the word is borrowed from Bengali, while "<E" marks the word as coming from English.

One problem that Mandi poses for the word collector is that a high proportion of its words are built up from two or more meaningful parts — from smaller morphemes. Words constructed from a half dozen meaningful pieces are by no means rare. Speakers construct words on the fly, just as English speakers construct phrases on the fly. It is no more possible to speak Mandi without constructing new words than it is to speak English without constructing new phrases. This means that Mandi words are not really the same sorts of units as English words. Obviously, a Mandi word list cannot include all the words that could be constructed. Such a list would never end. Ideally, one should list "lexical items" the conventionalized bits from which longer sequences are constructed. Such a list would have to include affixes that are never used as independent words, but it would also have to include many words that contain several morphemes, but whose overall meanings have become so conventionalized that they are not fully predictable from the meaning of their parts. Unfortunately, it is impossible for this non-native speaker to judge conventionality in a reliable and consistent way. I doubt very much if native speakers could do so either.

In fact, the list includes many words whose meanings are largely, or even completely, predictable from their parts. Native speakers should be able to construct many of these words without having heard someone else use them first. Such words need to be listed because they illustrate the processes of word formation and, in fact, even if native speakers could make them up, they do not have to. They have heard these words from other speakers, and they do not need to make them up fresh every time they are needed.

The available English to Garo dictionaries are not well designed for foreign learners. They tend to give Garo definitions of English words rather than Garo equivalents. This is useful for the Garo speakers for whom the dictionaries were written, but it is confusing for the English speaking user. The Garo to English dictionaries are much more useful, and three reliable ones have been published (Nengminza 1988, Marak 1946 and later, Holbrook 1998). All these report on the A˙A chik dialect, and all of them include many words not found in this volume. For many purposes, these dictionaries will be more useful than the list given here, especially for those primarily interested in the A˙A chik form of the language. This new list has several complimentary strengths.

1. First, since this list is organized by semantic categories, the reader can find the particular area of meaning in which he is interested and find a range of words that deal with the topic. Of course, the lists are not Page  ix"complete". It would be quite impossible to give complete lists because the borders between the categories are so porous. Still, an inspection of the categories will suggest the sorts of things that Mandis have found interesting enough to deserve names, and an inspection of the words in a single category will suggest the ways in which meanings are organized in one of the many and varied areas of life.

2. Second, this list focuses on a different dialect than the published dictionaries. Virtually all earlier published works on the Garo language have dealt with the dialect that was originally spoken in the far northeastern corner of the area of Garo settlement. This was the dialect spoken by the first educated Garo speakers, and it is the dialect upon which the written language is based. It is the dialect that, thanks to its educated speakers and the written language, has become a sort of de facto standard for educated Garos in India. Even in Bangladesh, it is recognized as having a special status and it is sometimes regarded as "better" than other ways of talking, though Bangladeshi Mandis rarely speak it with any ease, and rarely attempt to do so.

The dialect of the Modhupur Mandis is spoken near the southern extreme of the Garo area, about as far from the northeastern area as any Garos live. The Modhupur dialect and "standard" A˙A chik are also probably about as different as any two dialects of Garo can be. (This does not include Ruga and A˙ tong which must count as separate languages.) With a bit of good will, even A˙A chik and the dialect of Modhupur are mutually intelligible with no more than an occasional misunderstanding. A comparison between the vocabulary represented here and that found in other published accounts of Garo will give some idea of the extent of variability. In order to suggest how the dialects differ, I have included a few distinctively A˙A chik words, but I have made no attempt to be systematic or complete. These A˙A chik words, and the varied pronunciations that are used, are often understood by Bangladesh Mandis, even though they are rarely used.

3. Third, I give more attention to variability than the published dictionaries. While Garo is not hugely variable from one region to another, other kinds of variability are great. Speakers within what seems to be a single culturally homogeneous village may pronounce some words in several different ways. It is simply not possible to list all the alternatives, so Garo dictionaries, like dictionaries of other languages, give an impression of greater homogeneity than the language really has. As with many other languages, the process of developing a writing system for Garo has tended to bring some standardization. Some forms have found their way into the dictionaries while others have not. The forms that get written tend to be seen as, in some way, "correct". Like any written list, mine fails to convey the full variability of the language, but I have made more effort than others Page  xto record alternative forms. This should give some idea of the range of variation that is to be found in the language.

4. Finally, I have included more affixes in my list than can be found in the published dictionaries. Garo dialects have many hundreds of affixes. They form a crucial component of the words. Since new words can be freely constructed by using these affixes, they are as much building blocks of larger constructions as are the words of English. They deserve their place in a list of the language's lexical items. Affixes such as -no-a 'future', that need to be suffixed to something else, are shown with an initial hyphen. Prefixes, such as the na˙-, which is never used alone but that serves as the first syllable of many names for varieties of fish, are shown with a final hyphen. Affixes, such as -ing- 'progressive', that occur in the middle of a word, and that need something both before and after, are shown with hyphens at both ends. Most of the bits that need two hyphens are adverbial affixes.

Many of the conventions used in Volume I are used here as well. I use the same spelling conventions for transcribing the language. "Mandi" continues to stand for the dialect of Garo spoken in Bangladesh and particularly for that of the Modhupur region. "A˙A chik" refers to the dialect orriginally based on that of the northeast corner of the Garo Hills that served as the basis of the written language and that has influenced the dialect of educated Garos in India. Other conventions used in this Volume will be found in the table on the next page.