Munda language
This article has been extracted from |
NOTE: While reading please keep in mind that all articles in this series have been scanned from a book. Therefore, paragraphs might have got rearranged or omitted and/ or footnotes might have got inserted into the main text of the article, interrupting the flow. Readers who spot these errors might like to correct (and/ or update) them in a Part II of this article. Secondly, kindly ignore all references to page numbers, because they refer to the physical, printed book.
The main article in The Linguistic Survey Of India
Kherwari
We pass to Central India, where we find the Munda languages occupying a strong position. The principal of these, Kherwari, with numerous dialects, has its head-quarters at the north-eastern end of the plateau of Central India, but has spread into, or left survivors in, the plains at its foot. It has many dialects, of which the best known are Santali and Mundari. At the other, the north-west, end of the plateau, in the western Districts of the Central Provinces and in Mewar, we find another Munda language, Kurku,1 which is said to have two dialects, -- Muwasi and Nahali, but, as stated above (p. 28), the latter is much mixed with other forms of speech and is on the verge of disappearing altogether. The other Munda languages are less important. They are spoken in the neighbourhood of Kherwari or to its South. The principal are Kharia, Juang, Savara, and Gadaba, and they are all more or less mixed forms of speech. Kharia is mostly spoken in the Ranchi District of Chota Nagpur, and has all the characteristics of a language that is dying out and is being superseded by an Aryan form of speech. Aryan principles pervade its grammatical structure and its vocabulary, and it is no longer a typical Munda language. It has been compared to a palimpsest, the original writing on which can only with difficulty be recognized. Juang is very similar. It is spoken by the Juangs or Patuas of the States of Keonjhar and Dhenkanal in Orissa. These people are probably the lowest in the scale of civilization of all the Munda tribes. Till quite recently the women of the tribe did not even sew fig-leaves together to make themselves aprons. A bunch of leaves tied on in front and another behind was all that was claimed by the most exacting demands of fashion, and this costume was ` renewed as occasion required, when the fair wearer went to fetch cattle from the wood which provided her millinery.' Attempts have been made to introduce the wearing of loin-cloths, but I know not with what success. The most southern forms of Munda speech are those spoken by the Savaras and the Gadabas of North-East Madras. The former have been identifed. with the Suari of Pliny and the Sabarae of Ptolemy. A wild tribe of the same name is mentioned in Sanskrit literature, even so far back as late Vedic times, as inhabiting the Deccan, so that the name, at least, can boast of great antiquity. Their language, is of considerable interest, and since it was discussed in Volume IV of the Survey a series of excellent Readers in it have been prepared by Mr. Ramamurti for the Madras
The languages of the Munda Branch must once have been spoken over a much greater area of India than their present habitat. In the South, and to a certain extent in Chota Nagpur, they have been superseded by Dravidian forms of speech, and in the North by Aryan or Tibeto-Burman tongues. In each case, however, they have left their mark. As for the Dravidian languages, it is very probable that the rules for the harmonic sequence of vowels, which form so prominent a feature of Telugu are due to their influeuce,1 and, to the North of Chota Nagpur, the extraordinary complexity of the verbal conjugation of the Aryan Bihari is equally probably due to the same cause.2 Another interesting point is that Munda is numeration is vigesimal. The speakers count by twenties, not by tens as we and other Europeans do. But among the peasantry of Northern India vigesimal counting is quite usual. Instead of saying ` fifty,' they say ` two score and ten, ' instead of ` sixty ' they say ` three score,' and so on. This might be a case of mere coincidence, but that it is really an old Munda survival is shown by the fact that kuri , the word used all over Northern India for ` a score ', is almost certainly a word of Munda origin. But it is in the Himalaya that these Munda survivals are most apparent. At the present day, the Mundas have themselves survived as a recognized people only in the wild hill-country of Central India, and it is in accordance with this that they should also have survived for a longer time in the forests of the Himalaya than on the Aryanized plains of Northern India. In the Himalaya, from North-East Assam to the North-East Panjab, the great mass of the inhabitants speaks various forms of Tibeto-Burman tongues. Most of these are quite pure of their kind and possess all the peculiarities proper to that form of speech. But between Darjiling, north of Bengal, and Kanawar, north of Simla in the Panjab, there is a series of scattered tribes speaking languages called in the Survey ` Complex Pronominalized.' Most of them belong to the group called by Hodgson ` Kiranti ', but there are also others not mentioned by him. These languages are all Tibeto-Burman, or belong to some group closely allied to the Tibeto-Burman, but through them all there runs a peculiar strain which it is impossible not to recognize as Munda, once attention is drawn to it.3 These Complex Pronominalized languages are many in number, and will be further dealt with when we come to the consideration of the Tibeto-Burman languages. Suffice it here to say that the most western is probably Kanawari, spoken in the Simla Hills, though there are doubtful cases even further west.
The Munda languages were first recognized as a separate group, distinct from the Dravidian, in the year 1854 by the late Professor Max Muller in his famous ` Letter to Chevalier Bunsen on the Classification of the Turanian Languages,' and received its name ` Munda' from him4. As stated on page 14, in the comity of scholarship it has ever been an established rule that the first discoverer of any fact, whether it be a newly described flower, a newly described mineral, or a newly described group of languages, should have the right to give it its name, and that that name should be employed by other students unless and until it has been proved to be entirely false and misleading. Unfortunately this comity was not observed in the present case. Twelve years later, Sir George Campbell, no doubt unwittingly, ignored the name already given by Mix Muller, and proposed to call these languages ` Kolarian'1 because, as he imagined, the word ` Kol,' -- a common tribal name of the Munda people, -- was derived from an older form ` Kolar,' which he apparently connected with the Kolar District of Mysore in Southern India, and looked upon as identical with the Kanarese word kallar meaning ` thief.' There is absolutely no foundation for this supposition, and this name ` Kolarian ' is not only based upon a fantastic error, but is, in itself, objectionable as seeming to suggest a connexion with the word ` Aryan ' which does not exist.
Nearly every word of the above applies with equal force to the Munda languages.
Among other characteristics of the Munda languages we may mention the following. As in the Indo-Chinese languages, final consonants are often checked, or pronounced without the offglide, thus forming what is often called by Chinese scholars the ` abrupt' or ` entering tone.' Such consonants are as characteristic of Cantonese as they are of Munda, and are common, so far as I am aware in all the languages of the Mon-Khmer branch of Austro-Asiatic speech.1 Although masculine and feminine nouns are distinguished, there are only two real genders, one for all animate and the other for all inanimate objects. Nouns have three numbers, a singular, a dual, and a plural, the dual and plural numbers being indicated by suffixing the dual or plural, respectively, of the third personal pronoun to the noun. Short forms of all the personal pronouns are freely used, in each case as verbal suffixes. The dual and plural of the first personal pronoun have each two forms, one including the person addressed, and the other excluding him. If, when giving orders to your cook, you say, 'we shall dine at half past seven', you must be careful to use ale for ` we,' not abon ; or else you will invite your servant also to the meal, which might give rise to awkwardness. As in many other eastern languages, participial formations are used instead of relative pronouns. ` The deer which you bought yesterday' would be rendered ` the yesterday deer bought by you.' Roots are modified in meaning not only by suffixes, but also by infixes, as in da-pa-l mentioned above. The logical form of a Munda sentence is altogether different from that of Aryan languages, and hence it is impossible to divide it into the parts of speech with which we are familiar, say, in English. The nearest thing that it has to what we call a verb merely calls up an idea, but is unable to make any assertion. The final assertion is made by one of the most characteristic features of Munda grammar, a particle known as ` the categorical a .' By its form, the sentence first unites the represented ideas into a mental picture, and then, by a further effort, affirms its reality. In English we say " John came." A Santali would first call up a picture of John having come, and then, by adding the categorical a , would assert that this picture was a fact. Hence this a is not used in sentences that do not contain a categorical assertion, e.g. those which in English would contain a verb in the subjunctive or optative mood. Munda, with what is really better logic, relegates subjunctive and, relative to what may be called the incomplete verb in company with what are with us participles, gerunds, and infinitives, and forms the only complete and real verb by the addition of the categorical a .
As in the case of several other civilised and semi-civilised tribes, the names which we give to many Munda tribes are not those by which their members call themselves, but those which we have adopted from their Aryan-speaking neighbours. Most of the tribes simply call themselves ` men', the same word with dialectic variations, Ko1, Kora, Kur-ku (merely the plural of Kur), Har, Hara-ko(another plural), or Ho, being used nearly universally. The Indian Aryans have adopted in one case the word ` Kol ' as a sort of generic term for any of these non-Aryan tribes, and have identified the word with a similarly spelt Sanskrit term signifying ` pig,' a piece of etymology which, though hardly in accordance with the ideas of European science, is infinitely comforting to those that apply it. The Raj of these Kols is a subject of legend over large tracts of the south side of the Gangetic valley, where not one sentence of Munda origin has been heard for generations. The name is perhaps at the bottom of our word ` coolie,' and of the names of one or more important castes which would indignantly deny their Munda origin.
Other references to this language in The Linguistic Survey Of India
NOTE: The following are not likely to be complete paragraphs. They are excerpts obtained through an Internet search.
shadows all the rest, -- that of Brian Houghton Hodgson, -- as the author of an article on the Language, Literature, and Religion of the Bauddhas of Nepal and Bhot (Tibet). This was followed by a long series of papers on the zoology and ethnology of Nepal, but, nineteen years afterwards, in 1847 (Journal A. S. B. Vol. XVI), he resumes his philological enquiries, with a Comparative Vocabulary of the Sub-Himalayan dialects. Then followed a number of important papers, still classics, and still full of varied and accurate information regarding nearly every non-Aryan language of India and the neighbouring countries. Space will not allow me to give even a dry catalogue of the subjects which he adorned. Suffice it to say here that he gave comparative vocabularies of nearly all the Indo-Chinese languages spoken in India and the neighbouring countries, and of the Munda and of the Dravidian forms of speech. These he compared with many languages of Central Asia in the search of one common origin for the whole. So far as I am aware, he was the first Englishman to use the term ` Dravidian' for the languages of Central and Southern India, but he included under that term not only the Dravidian languages proper, but also those of an altogether different family, -- the Munda. It is true that he failed to establish his favourite theory of a common origin for all the languages explored by him, -- that is a matter still under inquiry: and on
the first time, the existence of the Munda1 family of languages as an independent body of speech, apart from the Draavidian, and gave it a name. Two years later, in 1856, appeared what has ever since been the foundation of research into the tongues of Southern India, Bishop Caldwell's ` Com
day, amongst whom special attention must he dramn to Pater W. Schmidt's brilliant work on ` Die Mon-Khmer-Volker' (1906). Pater Schmidt has here proved not only that the Mon-Khmer languages form a link between the Munda languages of India proper and the languages of Indonesia, -- grouping the first two, with Khasi and some other minor forms of speech, under the
Erskine Perry, then Chief Justice of Bombay and President of the Society, published his paper ` On the Geographical Distribution of the principal Languages of India.' He divided the languages of India into two great classes, -- ' the language of the intruding Arians, or Sanskritoid, in the North, and the language of a civilized race in the South of India, represented by its most cultivated branch, the Tamil.' The former he reckoned as seven in number, viz., Hindi, Kashmiri, Bengali, Gujarati, Marathi, Konkani, and Oriya, with ten dialects. Panjabi, Lahnda (called by him Multani), Sindhi, and Marwari he looked upon as all dialects of Hindi. Maithili he classed as a dialect of Bengali. Since he wrote, it will be seen that many of the forms of speech that he looked upon as dialects have been raised to the dignity of being recognized as independent languages. The Southern languages he called ` Turanian or Tamiloid.' He did not seem to be aware of- the term ` Dravidian' which was first used simultaneously in 1856 both by Hodgson and. by Caldwell. Perry mentioned Telugu, Kanarese, Tamil, Malayalam, Tulu, and with a query) Gondi. He gave brief descriptive accounts of the general characteristics of each language, and carefully indicated the habitat of each, the whole being illustrated by an excellent language map. It will be observed that he altogether ignored the Indo-Chinese languages, and that he made no mention of the Munda languages, which were not identified by Max Muller till the following year. While Perry confined himself to the geographical distribution of the Indian languages, anther Bombay scholar was studing the interaction between Indo-Aryan and Dravidian languages. The same volume of the Journal of the Bombay Branch of the R. A. S. contains J. Stevenson's Comparative Vocabulary of the Non-
adoption of theories as to relationship.1 So much could not be helped ; but beyond this every effort has been made to prevent the Survey becoming an encyclopaedia of Indian philological science. That will, we may hope, follow when scholars more competent than the present writer have had time to digest the immense mass of ordered facts now placed at their disposal. Indeed, a beginning has already been made. Reference has already been made to Pater Schmidt's discoveries regarding the Austric languages, and it has been a legitimate source of gratification to me to observe the free use of the Survey which has been made by Monsieur Jules Bloch in his researches into Marathi, by Professor Turner and Professor Suniti Kumar Chatterji in their important studies in Gujarati and Bengali, and by Dr. Paul Tedesco in his luminous essays on the history of Aryan languages. One interesting result of Pater Schmidt's inquiries may here be added, as it has a direct connexion with the Survey. The Munda languages, as we know, belong to Chota Nagpur and the centre of India. It is also a familiar fact that the languages spoken in the Himalaya, far to the north of these Munda languages, are Tibeto-Burman in character. But even here the Survey shows us that there is a line of peculiar forms of speech, extending from Darjiling to the Panjab, that show evident traces of a previously existing language of the Munda family, which has been overlaid, so to speak, by the Tibeto-Burman of the later immigrants. There is thus evidence to show the existence, at some very ancient time, of a common language of which traces are still visible from Kanawar in the Panjab down through Further India and across the Pacific Ocean as far as Easter Island and New Zealand. Philology is not to be confounded with Ethnology, and here we may leave these interesting facts in the hands of ethnologists for further examination.
languages and dialects are recorded. This is the number found in the list given in Appendix I, in which the figures for each are compared with those of the Census of 1921. But in this enumeration there is. a good dial of double counting, as each language and each dia1ect is there given a separate number. A better idea of the results will be gained from the consideration that the Census of 1921 records 190, and the Survey records 179 languages, as distinct from dialects. When counting dialects, it must be borne in mind that, in order to make the total for the dialects tally with the number of the speakers of the language of which they form the members, it has been necessary to count the standard form of the language as one of the dialects. There are also, inevitably, cases in which a language has been returned, but its dialects not mentioned. For instance, the Khasi language (No. 8 in the list) and its dialects are arranged as follows : -- Khasi, Standard, Lyng-ngam, Synteng, War, Unspecified . Here, if we count Kbasi in the list of languages, we must omit ` Standard' and ` Unspecified' in counting our list of dialects and languages, or we shall be recording the same form of speech twice, or perhaps three times, over. Hence, in the above example, we can count only three dialects as additional to the standard Khasi language. On this principle, the 1921 Census has recorded 49 dialects in addition to the general language-names. The Survey, on the other hand, has recorded no less than 544 dialect-names in addition to the standard and unspecified forms of the 179 languages. The various forms of speech noted are therefore 237 (188 + 49) in the Census, and 723 (179 + 544) in the Survey. Each of these 723 is described in the Survey, in most cases with more or less complete grammatical accounts. A summary of the dotails2 of these figures is as follows : -- | SURVEY FIGURES. Mon-Khmer Branch | 1 Munda Branch
of tribes which have in historic times abandoned one language and taken to another. A striking example is afforded by the tribe of Nahals in the Central Provinces. These people appear to have originally spoken a Munda language
akin to Kurku. It came under Dravidian influence, and has become a mixed from of speech, half Munda and half Dravidian. This, in its turn, has fallen under the spell of Aryan tongues, and is now in a fair way to becoming an Aryan language.1 If we were to judge by language, a hundred years ago we should have called the tribe Munda. Ten years ago it was quite possible to claim it as Dravidian, and fifty years hence it would probably be described as an Aryan caste. The ` unholy alliance ' between the two sciences has long been condemned, and has now fallen into disrepute, and I have hence, in the following pages, refrained so far as was possible from discussing questions of racial origin. When I have done so, it has only been to being forward theories regarding the origin of nationalities which have been previously suggested by professed ethnologists, and to attempt to throw light on them when they are confirmed by philology. In one case only is it sometimes permissible to draw inferences as to race from the facts presented. by language. When we find a small tribe clinging to a dying language, surrounded by a dominant language which has superseded the neighbouring forms of speech, and which is superseding its tongue too, we are generally entitled to assume that the dying language is the original tribal one, and that it gives a clue to the latter's racial affinities. Take as an example the Malto spoken by the hillmen of Rajmahal . This language is decadent, and is surrounded by others which are superseding it. Even if we did not know it on other grounds, we should be justified in asserting that its speakers are Dravidian, because their tongue falls within that family. But even this relaxation of the general rule, which was first suggested. to me by Sir Herbert Risley, must, as the case of the Nahals shows, be exercised. with caution. The Nahals are probably Munda by race, but their present speech is almost Dravidian. Their decadent language is a twofold palimpsest. It first began to be superseded by Dravidian, and now it is being superseded) by Aryan. A careless application of Sir Herbert's theory would compel us at the present day to assume that the tribe was of mixed Munda and Dravidian origin. With a dominant language we can make no such relaxation. In India, the Indo-Aryan languages, -- the tongues of civilization and of the caste system with all the power and superiority which that system confers upon those who live under its sway, -- are continually superseding what may, for shortness, be called the aboriginal languages such as those belonging to the Dravidian, the Munda, and the Tibeto-Burman families. We cannot say that a Tibeto-Burmau Koch or a Dravidian Gond is an Indo-Aryan, because he speaks, as he often does, an Indo-Aryan language. The language of the Brahuis of Baluchistan is Dravidian, but manly of the tribe speak the Eranian Balochi in their own homes, and, on the other side of India, some of the tribe of Kharias speak a Munda, others a Dravidian language, and others, again, the Indo-Aryan Bengali. It may be added that nowhere do we see the reverse process of a non-Aryan language superseding an Aryan. It is even rare for one Aryan-speaking nationality to abandon its language in favour of another Aryan tongue. We continually end tracts of country on the borderland between two languages, which are inhabited by both communities, living side by side and each speaking its own language. In some localities, such as the District of Malda in Bengal, the Survey actually found villages in which three languages were spoken, and in which the various tribes had evolved a kind of lingua franca to facilitate intercommunication, while each adhered to its own tongue for conversation amongst its fellows. The only exception to this general rule about the non-interdhangeability of Indo-Aryan languages
is first the great Mon-Khmer Branch spoken in Further India, of which we have three representatives in Burma, in the shape of Mon, an ancient literary language now spoken in Thaton and Amherst, and Palaung and Wa, less civilized languages spoken in Upper Burma. Khmer and a number of other minor forms of speech belong to IndoChina, beyond the Burma frontier. Among the latter, mention may be made of two languages spoken by wild tribes of Malacca, the Sakei and the Semang. Like Khmer these are spoken outside the limits of British India. Nicobarese also belongs to this branch, and seems to form a connecting link between the Munda languages and Mon.
the Khasi and Jaintia Hills of Assam. This was fully dealt with in the Survey. Its standard dialect has been often described, and moreover possesses a small literature with which it has been endowed by the local missionaries. Khasi is more or less isolated alike from its cousins of Burma and from those of India, and has struck out on somewhat independent lines apart from Mon, Nicobarese, and Munda, which are mutually more closely connected than any of them is with Khasi. With its three dialects of Lyng-ngam, Synteng, and War, in addition to the standard form of speech, Khasi forms an island of Mon-Khmer speech, left untouched in the midst of an ocean of Tibeto-Burman languages. Logan was the first to suggest, and Kuhn subsequently showed conc1usively, that it and the Mon languages belong to a common stock. The resemblances in the vocabularies of Khasi and of the dialects of the PalaungWa group settle the question. But the resemblance is not only one of vocabulary. The construction of the Mon and of the Khasi sentence is the same. The various component parts are put in the same order, and the order of thought of the speakers is thus shown to be the same. Like Mon and other members of the branch, and unlike the other Indo-Chinese languages by which it is surrounded, Khasl has no tones.1 On the other
In such characteristics the dialects in question have struck out lines of their own, in entire disagreement with Tibeto-Burman, or even Tibeto-Chinese, principles. They have accordingly become modified in their whole structure. It is difficult to help inferring that this state of affairs must be due to the existence of an old heterogeneous substratum of the population, which has exercised an influenee on the language. That o1d population must then have spoken dialects belonging to a different linguistic family, and the general modification of the inner structure of the actual forms of speech must be due to the fact that the leading principles of those old dialects have been engrafted on the languages of the tribes in question. Now it will be observed that all these features in which the Himalayan dialects differ from other Tibeto-Burman languages are in thorough agreement with the principles prevailing in the Munda forms of speech. It therefore seems probable that Mundas, or tribes speaking a language connected with those now in use among the Mundas, have once lived in the Himalaya and have left their stamp on the dialects there spoken at the present day.
The influence of the ancient language of the Munda type is not so prominent in these languages as in those of the pronominalized group. There are nevertheless distinct traces of its previous importance, and we may assume with considerable probability that here we have a case of the old influence receding before that of Tibetan and of the Bodo languages spoken immediately to the East. We appear to have a clear example of this in Sunwar. In Hodgson's days it was a pronominalized language, but, if the specimens received for the Survey are to be trusted, it is so no longer. Hodgson's Essay was written in 1847, so that, allowing for the date when the specimens for the Survey were received, this change took place in little more than half a century. As we know how rapidly Tibeto-Burman languages which have no literature to act as a conservative influence do change, this short period need not surprise us, and it is pretty certain that in all these languages the Munda characteristics were much stronger two or three centuries ago than they are now. On the other hand we also see in these non-pronominalized languages links connecting them with the Bodo Group. Whether they are naturally inherent in the languages or have been borrowed from the neighbouring languages we do not know, but, either way, it is the presence of these links which cause the Himalayan languages to form the western limb of the letter Y alluded to on page 53.
In the Bronominalized group the influence of the ancient Munda language is far more apparent. In all of them we notice the characteristic idiom of suffixing personal pronouns to the verb to indicate not only the subject but also, often, the direct and indirect objects. When a Limbu wishes to say ` I strike him,' he turns both the ` I' and the ` him' into suffixes added to the verb. ` Strike' is hip , ` him' is -tu , and ` I' is -ng , so he says hiptung , which it will be remembered is exactly parallel to the Santali example given on page 37. Some of the languages of this group follow the Munda system of courting the higher numbers in twenties. Only two follow the Tibetan system of counting by tens, and the rest have embarrassed comparative philology by borrowing the Indo-Aryan numerals. In Tibetan and the languages allied to it there is a complicated system for expressing pronouns. But the various forms are due to the exigencies of etiquette, and each implies a different degree of politeness, just as in many other oriental languages we hear such expressions as ` this poor slave' used instead of an uncompromisingly egotistical ` I.'
But in these pronominalized languages, though there is great variation of pronominal forms, this is based on an altogether different principle. Exactly as in Munda, there are three forms indicating number, -- a singular, a dual, and a plural, -- for each person, and for the first person we have even greater diversity, there being separate duals for ` I and thou,' and ` I and he,' and plurals for ` I and you,' and ` I and they.' In some of the Western dialects we even find what might almost be called instances of borrowing of Munda words, and a relic of Munda or Mon-Khmer pronunciation in the checked final consonants which have been described.
We know more about the Western Group of the pronominalized languages, as they are all spoken in British India. They possess all the Munda characteristics that have become Aryanized, and have adopted the Aryan languages of their conquerors while they have retained their ethnic characteristics. Besides these, many millions of people inhabiting central anal southern India possessing the physical type classed by ethnologists as ` Dravidian ' are almost the only speakers of two other important families of speech, the Munda and the Dravidian proper. Owing to the fact that these languages are nearly all spoken by persons possessing the same physical type, many scholars have suggested a connexion between the two families of speech, but a detailed inquiry carried out by the Linguistic Survey shows that there is no foundation for such a theory. Whether we consider the phonetic systems, the methods of inflexion, or the vocabularies, the Dravidian have no connexion with the Munda languages. They differ in their sounds, in their modes of indicating gender, in their declensions of nouns, in their method of indicating the relationship of a verb to its objects, in their numeral systems, in their principles of conjugation, in their methods of indicating the negative, and in their vocabularies. The few points in which they agree are common to many languages scattered all over the world.
The above remarks conclude our survey of the Himalayan Tibeto-Burman dialects. As previously pointed out, the indications of the ancient Munda influence on these forms of speech is a matter of the greatest interest. It connects languages spoken in Lahul, Chamba, and Kanawar with the Munda languages of Central India, and, through them, with the Khasi spoken in Assam, and with the Mon-Khmer languages of Further India. These last lead us on to the tongues of Indonesia and Polynesia till we arrive at Easter Island. Roughly speaking, we find this Austric Family of languages extending from 80 east longitude to 110 west longitude, a total of 170 degrees longitude, or very nearly half way round the world. Excepting the Indo-European (which has in modern times spread from Europe to America) it is the most widely extended of any of the language families of the earth.
I have already alluded to the attempts made to prove a connexion with the Munda languages, and have explained how this cannot be considered to exist. Finally allusion may be made to comparisons with the Australian languages, and to suggestions of a possible connexion by laud between India and Australia in the times when the prehistoric Lemurian Continent is believed to have existed. That certain resemblances in language have been found cannot be denied, but, as yet, we cannot quote anything as proving that a linguistic connexionn is probable. All that we can say with our present knowledge is that it is not impossible. Up to a few years ago the knowledge of the Australian languages possessed by European scholars was very scanty. In 1919 Pater W. Schmidt1 succeeded in reducing order out of chaos, and in classifying the numerous cognate tongues spoken in that great island-continent. The next stage in the investigation will be to carry on the inquiry into New Guinea, and thence into India. This inquiry was actually begun under Pater Schmidt's auspices2 but was interrupted during the War, and up to the date of writing nothing has appeared on the subject. We can only, for the present, wait and hope that in the near future sufficient materials will be forthcoming to settle the question once for all.
guages of the north. The northern limit of this southern block of Dravidian languages may roughly be taken as the north-east corner of the district of Chanda in the Central Provinces. Thence, towards the Arabian Sea, the boundary runs south-west to Kolhapur, whence it follows the line of the Western Ghats to about a hundred miles below Goa, where it joins the sea. The boundary eastwards from Chanda is more irregular, the hill country being mainly Drawidian with here and there a Munda colony, and the plains Aryan. Kandh, which is found most to the north-east, is almost entirely surrounded by Aryan-speaking Oriyas. Besides this solid block of Dravidian-speaking country, there are islands of languages belonging to the family far to the north in the Central Provinces and Chota Nagpur, even up to the bank of the Ganges at Rajmahal. Most of these are rapidly falling under Aryan influences. Many of the speakers are adopting the Aryan caste system and with it broken forms of Aryan language, so that there are in this tract numbers of Dravidian tribes to whose identification philology can offer no assistance. Finally, in far off Baluchistan, there is Brahui, concerning which, as already stated, it is uncertain whether it is the advance guard or the rearguard of a Dravidian migration.
anything like the wonderful luxuriance of agglutinative suffixes which we have noticed as distinguishing the Munda family. They represent, in fact, a later stage of development, for, although still agglutinative, they exhibit the suffixes in a state in which they are beginning to be modified by euphonic onsiderations, dropping
In the Dravidian languages all nouns denoting inanimate substances and irrational beings are of the neuter gender. The distinction of male and female appears only in the pronouns of the third person, in adjectives formed by suffixing the pronominal terminations, and in the third person of the verb. In all other cases the distinction of gender is marked by separate words signifying ` male' and ` female.' Dravidian nouns are inflected, not by means of case terminations, but by means of suffixed postpositions and separable particles. Dravidian neuter nouns are rarely pluralized. The Dravidian dative ( ku , ki or ge ) bears no analogy to any case termination found in Sanskrit or other Indo-European languages, the resemblance to the Hindi ko being accidental. Dravidian languages use postpositions instead of prepositions. In Sanskrit adjectives are declined like substantives, while in Dravidian adjectives are incapable of declension. It is characteristic of Dravidian languages in contradistinction to Indo-European, that, wherever practicable, they use as adjectives the relative participles of verbs in preference to nouns of quality, or adjectives properly so called. A peculiarity of the Dravidian dialects (shared however with Munda) is the existence of two pronouns of the first person plural, one inclusive of the person addressed, and the other exclusive. The Dravidian languages have no passive voice, this being expressed by verbs signifying ` to suffer,' etc. The Dravidian languages, unlike the Indo-European, prefer the use of continuative participles to conjugation. The Dravidian verbal system possesses a negative as well as an affirmative voice. It is a marked peculiarity of the Dravidian languages that they make use of relative participial nouns instead of phrases introduced by relative pronouns. These participles are formed from the various participles of the verb by the addition of a formative suffix. Thus, ` the person who came' is in Tamil literally ` the who-came.'
their south, scattered amid a number of Munda languages. we find the Dravidran Kurukh or, as it is often called, Orao. Still further north, on the Ganges bank, we find the closely related Malto spoken by the Maler of Rajmahal. According to their own traditions, the ancestors of the tribe speaking these two languages lived originally in the Carnatic, whence they moved north up
mentioned, the so-called ` Desya ', or ` Local ', words of the Indian grammarians. It included all words which the grammarians were unable to refer to Classical Sanskrit as their origin. Many such words were included in this group simply through the ignorance of the writers who catalogued them. Modern scholars can refer most of these to Sanskrit like any other Tadbhavas. A few others are words borrowed from Munda or Dravidian languages. The great majority are, however, words derived from dialects of the Primary Prakrit which were not that from which Classical Sanskrit has descended. They are thus true Tadbhavas, although not in the sense given to that word by Indian grammarians, in whose philosophy the existence of such ancient dialects was not dreamed of. These Desya words were local dialectic forms, and, as might be expected, are found most commonly in literary works hailing from countries like
centred mainly round what are known as the cerebral letters of the alphabet. These letters did not occur in the original Aryan (i.e. Indo-Eranian.) language, and, in Indo-Aryan languages, came into being on Indian soil. They are common in Dravidian, as well as in Munda, languages, and in them were certainly not borrowed from IndoAryan. The point in discussion was whether the Indo-Aryans borrowed them from the Dravidians or whether they did not. Neither contention was entirely correct. These letters occur with frequency in words of purely Aryan origin. It would be more accurate to say that in many cases the pronunciation of Aryan words became changed under the influence of the example of the surrounding non-Aryan tongues, whose speakers many times exceeded the Indo-Aryans in numbers. Analogy did the rest, save that a certain number of words (such, for instance, as names of things of which the Aryans had no previous experience in their Central Asian home, or words of very common occurrence and in everyday use) were directly borrowed1. This is borne out by the fact that, where we have reason to believe that Dravidian influence was least strong, the use of these cerebral letters is most fluctuating. Thus, in Assamese, although the distinction is maintained in writing, there is practically no distinction in pronunciation between the dental and the cerebral letters. It is probable, also, that in other cases the Dravidian languages have had. an indirect influence on the development of the vernaculars. When there were two or three ways of saying the same thing, the tendency would be to employ the idiom which was most like in sound to an expression meaning the same thing used by the 'surrounding non-Aryan tribes. Thus, in the Prakrit stage, there were many ways of expressing the dative. One of them consisted in suffixing the Aryan word kahu (derived from the Old Sanskrit krite ), and it had most chance of surviving, because it resembled the Dravidian dative suffix ku , or the old Dravidian suffix from which the modern ku is descended. And so, owing to the existence of the suffix ku , this Aryan suffix kahu did survive to the exclusion of other dative suffixes in some of the Indo-Aryan vernaculars, and now appears in Hindi under the form of the familiar ko . Other similar instances of this non-Aryan influence on the Aryan languages of India could easily be quoted. Two will since. In the progress of a word through the stage of the Secondary Prakrits, a medial hard consonant first became softened, and then disappeared. Thus the Old Sanskrit chalati , ` he goes,' first became chaladi , and then chalai . Some of the Secondary Prakrit dialects remained for a much longer period than others in the stage in which the softened consonant is still retained. Nay, this softened consonant has in some cases survived even in the modern vernaculars. Thus the Old Sanskrit soka , ` grief,' is soga , not soa , in Hindi. The occasional retention of this soft medial consonant can be explained by the influence and example of the Dravidian languages, in which it is a characteristic feature. In some Dardic languages, and in some Indo-Aryan languages of the Outer Circle, especially in Kashmiri, Sindhi, and Bihari, a final short i or u is not dropped, as is usual in the Inner languages, but is, so to speak, only half-pronounced, the mere colour, as it were, of the vowel being given to the finial consonant. Thus the Sanskrit murti , ` an image,' becomes murat in the Inner Hindi, but is pronounced marati in the Outer Bihari. This is also characteristic of Dravidian tongues.
The influence of Munda languages on the Indo-Aryan tongues is not so evident.
These languages appear to have been superseded on the Gangetic plain of India by Dravidian before the Aryans had occupied that tract, but a few ancient Munda, or Austro-Asiatic, words appear in Sanskrit. Such are the names of things like betel, cotton, cotton cloth, or bamboo arrows which were new to the invaders,1 or else geographical names taken over by them, such as Kosala, Tosala, Kalinga, Trilinga, and several others 2. At present the Munda languages are confined to the forest country south of the plain, although, as explained above3, traces of them can be recognizedd as surviving in the Tibeto-Burman languages of the Central Himalaya as far west as Kanawar in the Panjab. As another Munda survival in the Indo-Aryan languages we may note the occasional counting by scores. While the Indo-Aryan numeral system is essentially decimal, the word kori , probably itself a Munda word, is commonly used for ` score', and the uneducated people of the Ganges Valley use this in the formation of the higher numerals. Thus ` fifty-two' would be expressed by them as ` two-score twelve ', do kori barah . This counting by twenties is a Munda peculiarity. The Mundas were strongest in the eastern portion of the Gangetic plain, and apparently exercised another kind of influence on the eastern dialects of Bihari. Here the conjugation of the verb is much complicated by changes depending on the number and person of the object. The word, for instance, ` beating ' is represented by one form in ` I am beating you ', and by another in ` I am beating him '. These changes are Aryan in origin, and have parallels in the languages of north-western India, but the system is that of the Munda verb4.
