The Language of the Modhupur Mandi (Garo), Volume 1

Simple Sentences

(A)

This chapter surveys the central features of Garo grammar, focusing on the way simple sentences are constructed. It is intended to serve both as an introduction to the grammar of the language and as a guide to the chapters that follow. These later chapters will describe in more detail the words and suffixes of the language and their organization into sentences.

I start with a sentence that is very simple but that demonstrates a number of the most important features of the grammar. This sentence, like most of the other sentences that are given in this book, is characteristic of the Mandi dialect of Bangladesh, but A'chik sentences are not much different:

This sentence consists of two noun phrases followed by a verb. The first noun phrase, ang-ni pi'-sa 'my child', is the subject of the sentence. The second noun phrase, consisting of just one word, nok-o 'at the house', is a locative. To call something a "locative" simply means that it shows the location where something happens. -o is the locative case marker. Many Garo noun phrases finish with a case marker that shows what role that particular noun phrase plays in the sentence. The -o shows that nok 'house' has the locative case. When used together nok and -o mean something like 'at the house' or 'in the house'. Subject nouns are used without a casePage  92marker in Garo, so the subject of this sentence, pi'-sa 'child', does not have a case marker. Since most noun phrases that are not subjects do have case markers, the absence of a case marker is enough to suggest that pi'-sa is the subject. The -ni of ang-ni is the possessive, or "genitive", case marker and it acts very much like the possessive -'s of English. It is attached to ang- , which is a form of the first person singular pronoun 'I', and when ang- and -ni are joined together, they form a possessive pronoun that means 'my'.

The third constituent of the sentence, cha'-ja-ing-a , is the verb, and like most Garo verbs it comes at the end of the sentence. Garo verbs need to have at least two parts, the verb base and what can be called the "principal verb suffix". The verb base in the example is cha'- 'eat' and the principal verb suffix is -a . This is a kind of "neutral" tense but it is often used to show present time. Pi'-sa cha'-a would mean 'The child eats'. In addition to a verb base and a principal verb suffix, many verbs also have additional affixes that are inserted between the two obligatory parts. The verb in the example has two such affixes: -ja- 'not' and -ing- 'progressive'. The progressive gives a meaning much like English '-ing-', so cha'-ja-ing-a means 'is not eating'. (The fact that both English and Mandi have progressive affixes that are spelled -ing- is pure coincidence. Their pronunciation is actually quite different, since Mandi i before a consonant such as -ng sounds very different from English i . The A'chik form of the progressive is -eng rather than -ing .

This is a very simple sentence but it demonstrates several important features of Garo grammar. The verb usually comes last. Nouns and noun phrases precede the verb and their role in the sentence is often shown by a case marker that comes at the end of the noun phrase. Both verbs and noun phrases can be very much more complex than those in this simple example, of course, and both need to be described in more detail.

Verbs

(A)

The simplest Mandi verb (and indeed the simplest Mandi sentence) requires nothing except a verb base and a principal verb suffix, but verbs often have many more constituents than just these two. The best way to get a feeling for the way in which verbs can be built from smaller parts is to examine a list of verbs. Notice that each example given below has a verb base and a principal verb suffix. The verb base always comes first in the word. The principal verb suffix is often, though not always, last. The principal verb suffixes that are used in these examples are -a 'neutral', -no-a 'future', -bo 'imperative', and -jok . The meaning of -jok is a bit subtlePage  93and will be explained more fully later, but for now it can be understood to convey the sense that English shows by the perfect tense. This is the tense constructed with 'have', as in 'I have gone'. Sok-jok , then, can be translated as 'have arrived'. The meaning of all these principal verb suffixes will be explained more fully in the next chapter. A Garo verb can, all by itself, act as a full sentence without an explicit subject, but it can sound odd to translate verbs into English without a subject. Therefore the translations include '(she)'. The parentheses indicate that nothing in the Garo sentence explicitly indicates 'she', but that this meaning might be understood from the general context. Of course a different context could lead to the sentences being understood as having 'I' or 'they' or 'the men' or a great many other things as the subject. Abbreviations are: Neut 'neutral tense'; Prf 'perfect tense'; Fut 'future tense'; Imp 'imperative'; Que 'question particle'.

Page  94

These verbs include a few examples of all the main types of verb constituents. The verb bases in the examples are sok- 'arrive', kat- 'run', a-gan- 'speak, talk, say', and re'- 'move through space'. Only re'- would be likely to strike a learner as at all odd. This verb means, roughly, 'to move oneself in some unspecified direction'. Usually, though not always, it is used with either -ba- or -ang- . These affixes show the direction of movement with respect to the speaker. re'-ba- means 'move in this direction, come' and re'-ang- means 'move in that direction, go'.

-ba- and -ang- are examples of what will be called "adverbial affixes". Other adverbial affixes used in these sentences are -pil'- 'return, back', ku 'still, yet', -ja- 'negative', and -ing- 'progressive'. None of these are obligatory but any or all of them can be used together, except that, as opposites, it would make no sense to use -ang- and -ba- in the same verb. Adverbial affixes almost always occur in a fixed order. If all these were used in the same verb they would appear as -ba-pil'-ku-ja-ing- literally 'here-back-yet-not-progressive', but more naturally translated as 'not yet (be)-ing back here again'. These are some of the most common adverbial affixes but there are scores of others. Using them skillfully is an essential part of speaking Mandi well.

A number of "terminating" suffixes can follow the principal verb suffix but only two are included among the examples. -ma forms yes-no questions.Page  95These are questions that can be answered with either a 'yes' or a 'no'. -ma comes at the very end of the verb and thus at the very end of the sentence. The other terminating verb suffix used here is -kon 'perhaps, probably', a very common way of expressing doubt.

The examples illustrate two pervasive aspects of Mandi grammar and, indeed, of the grammar of all dialects of Garo. First, a high proportion of the meaningful chunks ("morphemes" as linguists call them) that are used to construct words are exactly one syllable long. Garo does have a few two-syllable morphemes, and a very few that are even longer, but not many. The two-syllable morphemes that are used in the examples are a-gan- 'speak', and -no-a 'future'. The individual syllables in these morphemes have no independent meaning of their own. Except for words borrowed from Bengali or English, a considerable majority of Garo morphemes have just one syllable.

Another characteristic of Garo that is illustrated by these examples is its remarkable regularity. Morphemes exhibit few changes as they join together to form words such as the verbs in these examples. Indeed, just two of the examples show any irregularity at all. In the next to the last example, the meaning 'back, return' is shown by -pil- , while earlier examples had -pil'- , with a raka. (This is the glottal stop, a catch made by quickly closing the vocal cords.) The variation between -pil- and -pil'- conforms to a very general Garo rule that rakas are always dropped from the second syllable of a word. The basic form of this morpheme should be regarded as -pil'- , with the raka shown by the raised dot, but it loses the raka whenever it forms the second syllable of a word. The final example shows a different kind of irregularity. -ja- is the regular sign of the negative in Garo. Ordinarily -a is the suffix that shows the neutral tense. When -ja- and -a would be expected to occur together in the same verb, however, the result is simply -ja rather than *-ja-a (the asterisk indicates that the form shown is not actually used in speech). In other situations, -ja- and the following suffix retain their full forms:

In summary: Verbs can include several constituents. First is the verb base. Second, there may be one or more adverbial affixes. Third, there must be a principal verb suffix. Fourth and last, there may be one or more terminating suffixes. The verb bases and principal verb suffixes, which formPage  96the obligatory components of verbs, are described in detail in Chapter 6, "Verbs". The adverbial affixes and the terminal suffixes, which are optional, are described in Chapter 7, "Optional Verb Affixes".

Noun Phrases

(A)

The only obligatory part of a Garo sentences is the verb. Neither a subject, object, nor any other type of noun phrase needs to be included in a sentence so long as the intended meaning is clear from the context. Many sentences, of course, do have one or more noun phrases and they can be used as subjects, objects, indirect objects, instrumentals (which show the instrument by which the action was accomplished), place locatives (to locate the action in space), time locatives (to locate the action in time), and in a number of other roles. The role of each noun phrase is generally made clear by means of a distinctive case marker that is suffixed to its final word. Since the role of each noun phrase is shown clearly by its case marker, the noun phrases can be placed in almost any order without obscuring the meaning, but normally they all come before the verb. The subject is likely to come before the object, but sometimes the object comes first instead. Differing orders of the noun phrases can imply slightly differing emphasis, but they do not change the fundamental meaning of the sentence. The dog bites the cat and The cat bites the dog mean very different things in English. In Garo, the direct object (the animal that is bitten) has to be clearly marked with the accusative case marker, -ko , and this allows it to be placed either before or after the subject, with no danger of confusing the biter from the one who is bitten. A-chak meng-gong-ko chik-a or Meng-gong-ko a-chak chik-a 'The dog ( a-chak ) bites the cat ( meng-gong )', but Meng-gong a-chak-ko chik-a or A-chak-ko meng-gong chik-a 'the cat bites the dog'.

Simple noun phrases can consist of nothing more than a noun and a case marker, and nouns that are the subject of a sentence do not even have a case marker, but noun phrases frequently have additional constituents. These can include any or all of the following:

Demonstratives. The most common demonstratives are i-a 'this' and u-a 'that'. These are frequent constituents of noun phases, and when present, they are always first in the noun phrase. Notice that, unlike English, Garo does not have obligatory articles (words such as the and a ), but when it is essential to indicate a definite meaning (the meaning conveyed by English the ), Garos use a demonstrative. As a result, demonstratives tend to be used somewhat more often in Garo than in English and they can often be naturally translated by 'the' rather than by 'this' or 'that': u-aPage  97a-chak 'that dog, the dog'. It is possible to use a demonstrative without a noun, and in this case it can be regarded as a demonstrative pronoun. The case markers are then attached directly to the demonstrative and the entire noun phrase is formed from a demonstrative and its case marker: i'-ko cha'-bo 'eat this'. -ko is the accusative (direct object) case marker. Demonstratives are described in more detail in Chapter 9, "Nominals".

Genitives (Possessives). Possessive nouns and possessive pronouns are always placed before the noun that stands for whatever is possessed. The case marker that shows possession for both nouns and pronouns is -ni : ang-ni a-chak 'my dog', u-a a-chak-ni ki'-me 'that dog's tail', nang'-ni jak 'your hand', nang'-ni ma-ma 'your maternal uncle'. Genitives are more fully described in Chapter 9, "Nominals".

Numerals. Garo numerals consist of two parts, first a classifier, then a number. Classifiers are chosen according to the type of object being counted. When counting people, the classifier is sak- or ak- . The classifier for animals is mang- . Still other classifiers are used for roundish objects, for thin flat objects, for long thin objects, and so forth: me'-chik sak-sa 'one woman', me'-a-sa sak-gin-i 'two men', ma'-su mang-gin-i 'two cows', a-chak mang-sa 'one dog'. In these examples, the numeral follows the noun that it modifies and that is the most common word order, but the numeral can precede the noun instead: sak-sa me'-chik 'one woman'. Garo has dozens of classifiers to choose among, and sometimes two or three different classifiers can be used for counting the same objects, the choice among them depending on the aspect of the object the speaker wants to emphasize.

Just as it is possible to use a demonstrative to convey a definite meaning, so is it possible to use a numeral formed from a classifier and sa 'one' to convey an indefinite meaning (the meaning shown in English by the indefinite article 'a'): Sak-sa me'-a-sa re'-ba-jok 'One man has come, a man has come'. It is also possible to use a numeral with no noun at all, and since the classifier gives considerable information about the thing being counted, Garos use numerals without nouns more often than English speakers do:. In sak-git-tam re'-ba-eng-a 'three people are coming', the classifier sak- shows that it is people, rather than animals, or busses, or anything else, that are coming. Since sak- is used only for people, no noun is needed to make that meaning explicit. Numbers and classifiers are described more fully in Chapter 10.

Modifying verbs. As will be explained more fully in the first section of the next chapter, the distinction between verbs and adjectives is less sharp in Garo than in English. Words that translate English adjectives can generally be used as intransitive verbs in Garo. At the same time, words that translate English verbs can be used to modify nouns in Garo. In otherPage  98words, the same words can be used in both ways, either as the verb of a sentence or as a modifier of a noun. When verbs are used as modifiers, a special type of principal verb suffix is used that puts the verb into a modifying form. The most explicit such suffix is -gip-a . From dal'-a 'to be big', for example, a modifying verb, dal'-gip-a can be formed: dal'-gip-a a-chak 'big dog'. Cha'-gip-a me'-chik 'the woman who eats', 'eating woman' is constructed from cha'-a 'eat' and me'-chik 'woman'. While modifying verbs are often placed before the noun they modify, they are sometimes placed afterwards instead. The suffix -a can form a modifying verb just as -gip-a can: me'-chik chon-a 'the small woman' is constructed from me'-chik 'woman' and chon-a 'to be small'.

Noun. The noun can be regarded as the central constituent of a noun phrase. It is the constituent that is modified by demonstratives, numbers, and verbal modifiers, and it is the noun's relationship to the rest of the sentence that is indicated by the final case marker. As in all languages, nouns form one of the largest classes of words, and Garo has thousands of them. Many nouns are single unanalyzable units (single morphemes) but even more are compounds, constructed from two or occasionally three or more parts, all of which are joined together into a single word. Typical nouns name body parts, plants and animals and their parts, natural objects, human-made artifacts, materials, kinship terms, and a very large number of more abstract phenomena. It is also possible to construct a noun from a verb by adding a "nominalizing suffix" to the verb, and a noun formed in this way can then be used as the central constituent of a noun phrase. Nouns and the formation of compounds will be considered in more detail in the first section of Chapter 8, "Nouns". The other constituents of noun phrases will be described in Chapter 9, "Nominals".

Pronouns. Like any language, Garo has personal pronouns with such meanings as 'I' and 'we'. Garo differs from English in making a clear distinction between 'you singular' and 'you plural', and it also makes a distinction between the 'we' that includes the person spoken to ("inclusive") and the 'we' that does not include the person spoken to' ("exclusive"). On the other hand Garo might be seen as less sexist than English since it makes no distinction between 'he' and 'she'. Garo does not have a pronoun that corresponds exactly to English 'it', but the demonstratives, i-a 'this', and u-a 'that' can be used in its place. Pronouns take case markers and they can act as noun phrases in much the same way as nouns can, but they rarely have modifiers of the kinds that are used with nouns. Pronouns have a few irregularities. Those in Table 5—1 that end with -a lose the -a when a case marker is added. This -a can be considered to be a nominative (subject) case marker that is used only with monosyllabic pronouns. The pronoun na'-a 'you singular' has a special form, nang'- , that is always used withPage  99case markers. ang-a 'I' (nominative), ang-ni 'my' (possessive), bi-ko 'him, her (accusative)', ching-o 'at us, by us' (locative). The pronouns of Mandi differ somewhat from those used in A'chik. The main pronouns of the two dialects are given in Table 5—1. Pronouns will be described more fully in Chapter 11, "Minor Parts of Speech".

Case Markers. Case markers show the relation of the noun or noun phrase to the rest of the sentence. The most important case markers are showin in Table 5—2. It is quite possible to have noun phrases with all five of these case markers in a single sentence. Four of the case markers need to be placed at the end of their noun phrases and they show the relation of their noun phrases to the rest of the sentence. The possessive case, shown by -ni , relates two nouns to each other, the possessor and the possessed, and both nouns occur within a single noun phrase.

The case system will be described in more detail in Chapter 8, "Nouns".

Postpositions. Instead of prepositions, Garo uses postpositions. The difference is merely that postpositions come at the end of a noun phrase, while prepositions come at the beginning. Garo postpositions follow the case marker of a noun phrase, but the line between postpositions and case markers is not entirely sharp. They can be distinguished easily whenever both a case marker and a postposition are used together, as they often are. Some postpositions regularly follow a particular case marker. The clearest examples are postpositions that follow the possessive case marker -ni . For example:

The postposition mik-kang is clearly derived from a noun that means 'face' but in this sentence it has the locative suffix -o and a locative meaning. In this usage, the locative -o is almost frozen to the noun, and mik-kang-o is used very much as we use 'in front of'. Front began as a body-part term just as mik-kang-o did, but both words have gained a specialized use to indicate a location.

Many, though by no means all, postpositions have the appearance of frozen, or partially frozen, locatives. This is hardly surprising since many postpositions indicate location in either time or space. Many postpositions, like mik-kang-o follow the possessive case marker, but others follow other case markers instead. Postpositions will be described in Chapter 9, "Nominals".

Minor Word Classes

(A)

In Garo, as in most languages, nouns and verbs are the two largest classes of words. Several other classes that occur as constituents of noun phrases, such as numerals and pronouns have already been discussed briefly. In addition to these, several other classes need to be recognized. Most of these are described more fully in Chapter 11, "Minor Parts of Speech".

Adverbs. The third most extensive word class in Garo, exceeded only by nouns and verbs, is that of adverbs. Unlike nouns and verbs these take no affixes. A large number of adverbs exhibit reduplication or partial reduplication, meaning that the same syllable or same sequence of syllables isPage  101repeated, sometimes with slight changes. Adverbs often occur immediately before the verb of a sentence, but they can also be placed earlier. As will become clear later, adverbs are closely related to verbs, and they are often derived from verbs. The following are typical:

Question words. Garo has words that are equivalent to the 'wh- questions words' of English such as who, when, where etc. These are used in sentences in ways that are similar to nouns, pronouns, or adverbs, but they ask questions that require a more precise answer than a simple 'yes' or 'no'. Most of them are used in much the same grammatical situations as nouns, and like other nouns they regularly take case markers.

Conjunctions. Garo makes less use of words that act as conjunctions than does English. In some cases, it uses suffixes where English would use conjunctions, and in some cases it simply places two words side by side without any overt conjunction at all. Nevertheless a few conjunctions and conjunction-like words are used: ar or a-ro 'and', ong'-ja-o-ba 'if not, or'.

Echo words. Garo has many hundreds, of "echo words" that follow and augment another more meaningful word. The echo word generally resembles the previous word by rhyming or by alliteration, hence the name "echo", but the two words are not identical. Some echoes can be used as independent words, without their partner, but many of them are used only as echoes and these add little meaning of their own except to give a certain weight or importance to their partners. The addition of the echo gives rhetorical color to the language, and this is one aspect of their language about which native speakers are clearly aware, and in which they take considerable pleasure. They are known to the speakers as ku'-pa me-li-a 'words that go well together'. Echo words are not restricted to just one part of speech: chat-chi tat-chi 'kinsmen', dong-su-sak dong-ru-rak 'restless', rang'-kan-ta chu-kan-ta 'angry', ma'-su mat-e 'cattle'. Most languages spoken in the region, Bengali and Assamese as well as languages belonging to the Tibeto-Burman family have echo words similar to those of Mandi. Indeed, Garo has even borrowed a fair number of echo words from Bengali: bi-sun ti-sun 'seeds', nas-ta nos-ta 'breakfast, snack'.

Interjections. Garo has a few words that are generally used in isolation rather than being incorporated into a sentence, and this means that they are not much involved in the syntax of the language. Many of these interjections express emotions: ai-au 'wow!', da'-nang 'oh, dear, how sad'.Page  102

What is a Word?

(B)

I have written this chapter as if it were obvious what a word is, but I have not yet justified the reasons for putting word breaks in some places, but not in others. In fact, it is not always obvious which sequences of meaningful morphemes should be considered parts of the same word and which should be divided among separate words. In a community where few people write their own language, the conventions of writing cannot tell us where word boundaries should fall, and the people themselves have no reason to think about the matter and no means of explaining word boundaries to an outsider who wants to write the language. Still, a number of criteria can help to define the boundaries of words.

First, the intonation, including rises and falls in pitch and changes in stress, tends to set off certain word-like chunks from one another. Second the order of meaningful elements is freer in some cases than in others. Other things being equal, where the order is rigid it is reasonable to consider the units to be a part of the same word, but where the order is free, it generally seems better to count the movable bits as separate words. This is not an entirely reliable criteria, however. Demonstratives, for example, always occur first in a noun phrase so they might be looked upon as prefixes, but other criteria suggest that they are much better considered to be separate words. In particular, demonstratives can be used alone, something that might not be expected of a prefix. The ability to be used alone, then, is the third and most reliable criterion for words. Any piece of the language that can be used by itself, with silence both before and after, but whose individual parts cannot be used alone, can be considered a word.

By this criteria everything from a verb base through the principal verb suffix has to be considered a single word, because none of the parts are capable of being used alone. This means that verbs can be very long. Even the terminating suffixes that follow the principal verb suffix must be considered to be a part of the preceding verb. It is true that the verb to which a terminating suffix is attached can be used alone and act as a word, but the suffix cannot. The suffix needs to cling to the verb so it lengthens the previous word and it does not contribute a new word to the sentence. Case markers are a part of the word that precedes them because case markers can never be used alone. Demonstratives are separate words because they can be used by themselves. Mostly, then, we can decide on word boundaries by paying attention to what can and cannot be used alone, but a few ambiguous cases remain even then, and a few arbitrary decisions seem unavoidable. Three constructions are particularly difficult.

Reduplicated Adverbs. When a single syllable is reduplicated it seems natural to consider the resulting pair to be a single word. The sylPage  103lables are said in rapid succession, nothing can intervene between them, and the syllables cannot, ordinarily, be used alone. When a three syllable sequence is reduplicated to form an adverb, however, it is not so clear that all six resulting syllables should count as a single word. Ka-sin-e-ka-sin-e 'slowly', is constructed from a form of the intransitive verb ka'-sin-a 'slow' and it is not impossible to use ka'-sin-e by itself. Nevertheless, consistency suggests that if single syllable reduplications are to be considered as single words so should three-syllable reduplications, and that is how I will write them.

Case Markers and Postpositions. I treat case markers as if they are part of the preceding word but I treat postpositions as if they are separate words. For the most part this seems satisfactory, but there are marginal cases where it is not entirely clear whether something should be considered to be a case marker or a postposition. It then becomes essentially arbitrary whether to call it a suffix or a separate word.

Incorporated Objects. As will be described more fully in Chapter 6, "Verbs", direct objects are used without a case marker if two conditions are fulfilled. First, the direct object must appear immediately before the verb and, second, the direct object must be indefinite. A definite direct object must have the accusative case marker -ko . Sometimes indefinite objects become so closely tied to the following verb that they seem to merge with it. Three verbs (or phrases) offer a particularly persuasive example: mik-su-a 'wash one's face', jak-su-a 'wash one's hands', ja'-su-a 'wash one's feet'. Mik- is a part of mik-kang 'face', while jak and ja'- mean, respectively, 'hands' and 'feet'. Except for these three verbs, su- appears only in su-gal-a 'wash (things)'. Su- is never used except in association with one of these other morphemes, but it obviously means 'wash'. It seems natural to consider jak-su- and the others to be compound verbs formed from what could originally have been a direct object but that has now become a part of the verb. There are other examples in which the object is less securely fixed to the verb and where it less easy to decide whether they form one word or two. Indeed, there is a gradation from independent objects, on the one hand, to forms like jak-su-a where the noun seems to be thoroughly incorporated into the verb. I find it impossible to draw a firm distinction between sequences that are so closely associated that the parts must be considered to form one word and others that form two words. There is an inevitable indeterminacy.

All of the classes of words and suffixes that have been introduced in this chapter will be described more fully in the chapters that follow. Chapter 6, "Verbs", has sections on verb bases and principal verb suffixes. Chapter 7, "Optional Verb Affixes", describes adverbial affixes and terminatingPage  104suffixes. Chapter 8, "Nouns", describes nouns bases and noun suffixes, including case markers. Chapter 9, "Nominals", describes noun phrases and their constituents: demonstratives, pronouns, question words, modifying verbs and postpositions. Chapter 10, "Numerals", describes both classifiers and numbers. Chapter 11, "Minor Parts of Speech", has sections on adverbs, adjectives, question words, conjunctions, echo words, and interjections. The final three chapters, 11, 12, and 13, deal with syntax. They describe the way that all the many classes of words can be joined together to form the phrases, clauses, and sentences that allow people to communicate.