1911 Encyclopædia Britannica/Bengali

From Wikisource
Jump to navigation Jump to search
17557691911 Encyclopædia Britannica, Volume 3 — BengaliGeorge Abraham Grierson

BENGALI, with Oriya and Assamese, three of the four forms of speech which compose the Eastern Group of the Indo-Aryan Languages (q.v.). This group includes all the Aryan languages spoken in India east of the longitude of Benares, and its members are the following:—

Number of speakers in
British India, 1901.
Bengali 44,624,048
Oriya  9,687,429
Assamese    1,350,846
Bihari 34,579,844
—————
Total 90,242,167

Of these Bihari is treated separately. In the present article we shall devote ourselves to the examination of Bengali together with the two other closely connected languages. The reader is throughout assumed to be in possession of the facts described under the heads Indo-Aryan Languages and Prakrit.

Bengali is spoken in the province of Bengal proper, i.e. in, and on both sides of the delta of the Ganges, and also in the Eastern Bengal portion of the province of Eastern Bengal and Assam. The name “Bengali” is an English word, derived from the English word “Bengal.” Natives call the Language.language Banga-Bhāṣā, or the language of Banga, i.e. “Bengal.” “Oṛiyā” is the native name for the language of Ōḍra or Orissa. Assamese, again an English word, is spoken in the Assam Valley. Its native name is Asamiyā, pronounced Ohåmiyā. All these languages have alphabets derived from early forms of the well-known Nagari character of northern India. That of Bengali dates from about the 11th century A.D. It is a cursive script which admits of considerable speed in writing. The Assamese alphabet is the same as that of Bengali, but has one additional character to represent the sound of w, which has to be expressed in the former language in a very awkward fashion. In Orissa, till lately, writing was done on a talipot palm-leaf, on which the letters were scratched with an iron stylus. In such circumstances straight lines would tend to split the leaf, and accordingly the alphabet received a peculiar curved appearance typical of it and of one or two other South Indian methods of writing.

The three languages are all the immediate descendants of Māgadhī Prakrit (see Prakrit), the headquarters of which were in south Behar, near the modern city of Patna. From here it spread in three lines—southwards, where it developed into Oriya; south-eastwards into Bengal proper, where it became Bengali; and eastwards, through Northern Bengal, into Assam, where it became Assamese. It thus appears that the language of Northern Bengal, though usually and conveniently treated as a dialect of Bengali, is not so in reality, but is a connecting link between Assamese and Bihari, the language of Behar. It is noteworthy that Northern Bengali and Assamese often agree in their grammar with Oriya, as against standard Bengali.

Omitting border forms of speech, Bengali, as a vernacular, has two main dialects, a western and an eastern, the former being the standard. The boundary-line between the two may be roughly put at the 89th degree of east longitude. The eastern dialect has many marked peculiarities, amongst which we may mention a tendency to disaspiration, the pronunciation of c as ts, of ch as s, and of j as z. In the northern part of the tract a medial r is often elided, and in the extreme east there is a broader pronunciation of the vowel a, like that in the English word “ball,” k is sounded like the ch in “loch,” and both c and ch are pronounced like s. The letter p is often sounded like w, and s like h, which again, when initial, is dropped. The distinction between cerebral and dental letters is lost, so that the words āţh and sāt are both pronounced ’āt. In the south-east, near Chittagong, corruption has gone even further, and the local dialect, which is practically a new language, is unintelligible to a man from Western Bengal. Throughout the eastern districts there is a strong tendency to epenthesis, e.g. kāli is pronounced kāĭl. A more important dialectic difference in Bengali is that between the literary speech and the vernacular. The literary vocabulary is highly Sanskritized, so much so that it is not understood by any native of Bengal who has not received special instruction in it. Its grammar preserves numerous archaic or pseudo-archaic forms, which are invariably contracted in the colloquial speech of even the most highly educated. For instance, “I do” is expressed in the literary dialect by karitēchi, but in the vernacular by kȯrcci or kȯcci. Oriya and Assamese may be said to have no dialects. There are a few local variations, but the standard form of speech, as a whole, is used everywhere in the respective tracts where the languages are spoken.

The three languages, being all children of a common parent, present many similar features. Oriya on the whole preserves the usual accentuation of the Indo-Aryan Languages (q.v.), seldom having the stress syllable farther back than the antepenultimate. Bengali, on the other hand, throws the accent as far back as possible, and this produces the contracted forms which we observe in the colloquial language, the first syllable of a word being strongly accented, and the rest being hurried over. Literary Bengali preserves the full form of the word, and in reading aloud this full form is adhered to. Assamese follows Bengali in its accentuation, but the language has never been the toy of euphuism. In its literature colloquial words are employed, and are written as they are pronounced colloquially.

In the following account of the three languages, Bengali, literary and colloquial, will be primarily dealt with, and then the points of difference between it and the other two will be described. Abbreviations used: A. = Assamese, Bg. = Bengali, O. = Oriya, Pr. = Prakrit, Mg. Pr. = Māgadhi Prakrit, Skr. = Sanskrit.

Vocabulary.—As already said, Literary Bengali abounds in tatsamas, or words borrowed in modern times from Sanskrit (see Indo-Aryan Languages), and these have also intruded themselves into the speech of the educated. So much has the false taste for these learned words obtained the mastery that, in the literary language, when a genuine Bengali or tadbhava word is used in literature it is frequently not put into writing, but the corresponding learned tatsama is written in its place, although the tadbhava is read. It is as though a French writer wrote sicca when he wished the word sèche to be pronounced. Similarly, the Bengali word for the goddess of Fortune is Lakkhī, but in books this is always written in the Skr. form Lakṣmī, although no Bengali would dream of saying anything but Lakkhī, even when reciting a purple passage ore rotunda. In fact, the vocal organs of most Bengalis are incapable of uttering the sound connoted by the letters Lakṣmī. The result is that the spelling of a Bengali word rarely represents its pronunciation. Oriya also borrows freely from Sanskrit, but there is no confusion between tatsamas and tadbhavas, as in Bengali. Assamese, on the other hand, is remarkably free from these parasites, its vocabulary being mainly tadbhava. In Eastern Bengal, where Mussulmans predominate, there is a free use of words borrowed from Arabic and Persian. Owing to geographical and historical circumstances, Oriya is to some extent infected by Telugu and Marathi idioms, while the Tibeto-Burman dialects and Ahom have left their marks upon Assamese.

Phonetics.—The three forms of speech agree in sounding the vowel a like the ȯ in “hot.” When writing phonetically, this sound is represented in the present article by ȯ. The pronunciation of this frequently recurring vowel gives a tone to the general sound of the languages which at once strikes a foreigner. In Bg. and A. a final vowel preceded by a single consonant is generally not pronounced. In Bg. this is only true for nouns, a final a being freely sounded in adjectives and verbs. In O., on the other hand, a final a is always pronounced. The sound of such a final a is in all three languages the same as that of the second o in “promote”; thus, the Bg. bara is pronounced bȯṛō. In Bg. a medial a sometimes has the sound of the first o in “promote,” as, for instance, in the word ban (bon), a forest. In A. and Eastern Bg. a medial a is often sounded like the a in “ball,” and is then transliterated ȧ. Ā has preserved as a rule its proper sound of a in “father.” The distinction between i and ī and between u and ū is everywhere lost in pronunciation, although in tatsama words the Sanskrit spelling is followed in literature. Thus, in Bg., the Skr. vyatīta is pronounced bétítō, with the accent on the first syllable. In A. the distinction between these long and short vowels is obliterated more than elsewhere, the reason being, as in Bg., the changes of pronunciation due to the shifting back of the accent. In O., the Skr. vowel is pronounced ru. Elsewhere it is ri. In O. the vowel ē is always long, but in Bg. it may be long or short, and in A. it is always short. The syllable ya preceded by a consonant has in Bg. the sound of a short e, so that vyakti is pronounced bekti. Moreover, in the same language the letter ē is often pronounced like the a in the German Mann, a sound here phonetically represented by a; thus, dēkha is sometimes pronounced dekhō, and sometimes dȧkhō or even dȧkō. The syllable , when following a consonant, also has this ȧ-sound, so that the English word “bank” is written byānk in Bengali characters. Ō in O. is always long. In Bg., when it has not got the accent it is shortened to the sound of the first o in “promote,” a sound which, as we have seen, is also sometimes taken by a medial a. In A. ō approaches the sound of u, and it actually becomes u when followed by i in the next syllable. The diphthongs āī (in tatsamas, i.e. the Skr. āi) and ai (in tadbhavas) are sounded like oi in “oil” in Bg. and O., while in A. they have the sound of oi in “going.” Similarly, in Bg. and O. the diphthongs āū and au are sounded like the au in the German Haus, but in A. like au in the French jaune, or the second o in “promote.” In colloquial Bg. the two syllables āi often have the sound of ē, as in khāitē (khētē), to eat.

In Eastern Bengal k has often the sound of ch in “loch.” In A. the consonants c and ch are both pronounced like s, and j and jh become zh (i.e. the s in “pleasure”) or (when final) z. The same tendency is observable in Bg., though it is usually considered vulgar. In parts of Eastern Bengal c is pronounced like ts. O. as a rule has the proper sound of these letters, but towards the south c and ch become ts and tsh when not followed by a palatal letter. The letters and ḍh, when medial, are pronounced as a strongly burred r, and are then transliterated and ṛh respectively. In A. and Eastern Bg. there is a strong tendency to pronounce both dentals and cerebrals as semi-cerebrals, as is done by the neighbouring Tibeto-Burmans. In A. and ṛh become r and rh respectively. In Bg. and A. has universally become n, but is properly pronounced in O. Y is usually pronounced as j, unless it is a merely euphonic bridge to avoid a hiatus between two vowels, as in kariyā for kari-ā. In A. the resultant j has the usual z-sound. When y is the final element of a conjunct consonant, in Bg. (except in the south-east) it is very faintly pronounced. In compensation the preceding member of the conjunct is doubled and the preceding vowel is shortened if possible, thus vākya becomes bȧkkyō. In A., while the y is usually preserved, an i is inserted before the conjunct, so that we have bāikyō. M and v when similarly situated are altogether elided in Bg., and this is also the case with v in A., in which language m under these circumstances becomes w; thus, smaraṇa becomes Bg. śśȯrȯn, A. swȯrȯn, and dvārā becomes Bg. and A. ddārā. R is generally pronounced correctly, except that when a member of a compound it is often not pronounced in colloquial Bg.; thus karma (kȯmmō). In North-eastern Bengali and in A. a medial r is commonly dropped; thus, Bg. karilām (kaïlām), A. kari (kaï).[1] The vulgar commonly confound n and l. O. has retained the old cerebral of Pr., which has disappeared in Bg. and A. The semi-vowel v (w) becomes b in Bg. and O., but retains its proper sound when medial in A. When Bg. wishes to represent a w, it has to write ōyā; thus, for chāwā it writes chāōyā. Similarly bārō, twelve, +yāri, friendship, when compounded together to mean “a collection of twelve friends,” is pronounced bārwāri. Bg. pronounces all uncompounded sibilants as if they were ś, like the English sh in “shin.” This was already the case in Mg. Pr. (see Prakrit). O., on the contrary, pronounces all three like the dental s in “sin,” while A. sounds them like a rough h, almost like the ch in “loch.” In Eastern Bg. s becomes frankly h, and is then often dropped. The compound kṣ is everywhere treated as if it were khy, In colloquial Bg. there is a tendency to disaspiration; thus dēkha is pronounced dȧkō and the Pr. hattha-, a hand, becomes hāt, not hāth. In Eastern Bg. there is a cockney tendency to drop h, so that we have ’āt, a hand, and kaïlām for kahilām, I said.

The above remarks show that O. has, on the whole, preserved the original sounds of the various letters better than Bg. or A.

Declension.—The distinction of gender has disappeared from all three languages. Sex is distinguished either by the use of qualifying terms, such as “male” or “female,” or by the employment of different words, as in the case of our “bull” and “cow.” The plural number is almost always denoted by the addition of some word meaning “many” or “collection” to the singular, although we sometimes find a true plural used in the case of nouns denoting human beings. Case was originally indicated by postpositions (see Indo-Aryan Languages), but in many instances these have been joined to the noun, so that they form one word with it. The following is the full declension of the singular of the word ghōṛā, a horse, in the three languages:—

Oriya. Bengali.  Assamese. 
 Nom.
 Acc.-Dat. 
 Instr.
 Abl.
 Gen.
 Loc.
ghōṛā
ghōṛāku
ghōṛārē
ghōṛāru
ghōṛāra
ghōṛārē
ghōṛā
ghōṛākē
ghōṛātē
ghōṛā-haïtē
ghōṛār
ghōṛāte or ghōrāy
ghōrā
ghōrāk
ghōrārē
ghōrāyē
ghōrār
ghōrāt

In Bg. and A. a noun often takes ē (e) in the nominative singular, when it is the subject of a transitive verb; thus Bg. bēdeē (from bēd) balē, the Veda says. In Bg. the nominative plural may, in the case of human beings, be formed by adding ā to the genitive singular; thus, santān, a son; gen. sing., santānēr; nom. plur., santānēra. The same is the case with the pronouns; thus āmār, of me; āmarā, we; tāhār, his; tāhārā, they. In Bihari (q.v.) the pronouns follow the same rule, and, as is explained under that head, the nominative plural is really an oblique form of the genitive. With this exception, the plural in all our three languages is either the same as the singular, or (when the idea of plurality has to be emphasized) is formed by the addition of nouns of multitude, such as gaṇ in Bg., māna in O., or bilāk in A.

We shall see that pronominal suffixes are freely used in all three languages in the conjugation of verbs. In the Outer languages of the north-west of India (for the list of these, see Indo-Aryan Languages) pronominal suffixes are also commonly added to nouns to signify possession. In most of the languages of the Eastern Group such pronominal suffixes added to nouns have fallen into disuse, but in A. they are still commonly employed with nouns of relationship; thus, bāp, a father; bopāi, my father; bāper, your father; bāpek, his father. Their retention in A. is no doubt due to the example of the neighbouring Tibeto-Burman languages, in which such pronominal prefixes are a common feature.

In all three languages the adjective does not change for gender, for number or for case.

The personal pronouns have at the present day lost their old nominatives, and have new nominatives formed from the oblique base. In the first and second persons the singulars have fallen into disuse in polite conversation, and the plurals are used honorifically for the singular, as in the case of the English “you” for “thou.” For the plural, new plurals are formed from the new singular (old plural) bases. In A., however, the old singular of the first person is retained, and the old plural plays its proper function. The Bg. pronouns are, mui (old), I; āmi (modern), I; tui (old), thou; tumi (modern), thou; , tini, he; ē, ini, this; ō, uni, that; , jini, who; , who?; ki, what?; kōn, what (adjective)?; kēha, anyone; kichu, anything; kōna, any. Most of the forms in the other languages closely follow these. The words in O. for “I” and “thou” are ambhē and tumbhē respectively. All these pronouns have plurals and oblique forms to which the case suffixes are added. These must be learnt from the grammars.

Conjugation.—It is in the conjugation of the verb that colloquial Bg. differs most from the literary dialect. There is no distinction in any of the three languages between singular and plural. Most of the old singular forms have survived in a non-honorific sense, but they are rarely employed in polite language except in the third person. The old plural forms are generally employed for the singular also. The usual base for the verb substantive, when employed as an auxiliary, is ach, be, derived from the Skr. ṛcchati. O., however, forms its past from the base tha (Skr. sthita-), and in South-western Bengal the base ţha, derived from the same original, is used for both present and past time. Only two of the old Skr.-Pr. tenses have survived in the finite verb, the simple present and the imperative. Thus, Bg. kari, I do; kar, do thou. The past is formed by adding pronominal suffixes to the old past participle in il (Skr. -illa-, a pleonastic suffix, see Prakrit), and the future by adding them to the old future participle in b (Skr. -tavya-, Pr. -avva-). Thus, Bg. karil-ām, done + by-me, I did; karib-a, it-is-to-be-done + by-me, I shall do. In Bg. there are two modern participles, a present (kar-itē) and a past (kar-iyā), and from these there are formed periphrastic tenses by suffixing auxiliary verbs. Thus, karite-chi (colloquial, kȯrci or kȯcci), I am doing; karitē-chilām (coll. korcilum or kȯccilum), I was doing; kariyā-chi (coll., korsi), I have done; kariyā-chilām (coll., korsilum), I had done. A past conditional is formed by adding pronominal suffixes to the present participle; thus, karitām (coll., kortum or kottum), (if) I had done. Similar tenses are formed in O. and A., but the periphrastic tenses are formed with verbal nouns and not with participles. Thus, O. karu-achĩ, A. kari-chõ, I am a-doing, I am doing. O. and A. have each a very complete series of gerunds or verbal nouns which are fully declined. In Bg. only one gerund, that of the genitive, is in common use.

In order to illustrate the conjugation of the verb, we here give that of the root kar, do, in its present, past and future tenses.

Oriyā. Literary
 Bengali. 
Colloquial
Bengali.
 Assamese. 
I do
Thou doest
He (non-honorific) does
He (honorific) does
I did
Thou didst
He (non-honorific) did
He (honorific) did
I shall do
Thou wilt do
He (non-honorific) will do 
He (honorific) will do
karñ
kara
karē
karanti
karilū
karila
karilā
karilē
karibū
kariba
kariba
karibē
kari
kara
karē
karen
karilām
karilē
karila
karilen
kariba
karibē
karibë
kariben
kȯri
kȯrō
kȯrē
kȯren
kȯllum, kȯrlum
kȯllē, kȯrlē
kȯllō, kȯrlō
kȯllen, kȯrlen
kȯrbō
kȯrbē
kȯrbē
kȯrben
karõ
karā
kare
kare
kårilõ
kårilā
kårile
kårile
kårim
kåribā
kåriba
kåriba

All the three languages have negative forms of the verb substantive, and A. has a complete negative conjugation for all verbs, made by prefixing the negative syllable na under certain euphonic rules.

Bengali Literature.—The oldest recognized writer in Bengali is the Vaishnava poet Caṇḍī Dās, who flourished about the end of the 14th or the beginning of the 15th century. His language does not differ much from the Bengali of to-day. He founded a school of poets who wrote Literature.hymns in honour of Krishna, many of whom, in later times, became connected with the religious revival instituted by Caitanya in the early part of the 16th century. In the 15th century Kāśī Rām translated the Mahābhārata, and Kr̥ttibās Ojhā the Rāmāyaṇa into the vernacular. The principal figure of the 17th century was Mukunda Rām who has left us two really admirable poems entitled Caṇḍī and Śrīmanta Saudāgar. Parts of the former have been translated by Professor Cowell into English verse, and both well deserve putting into an English dress. With Bhārat Candra, whose much admired but artificial Bidyā Sundar appeared in the 18th century, the list of old Bengali authors may be considered as closed. They wrote in genuine nervous Bengali, and the conspicuous success of many of them shows how baseless is the contention of some native writers of the present day that modern literary Bengali needs the help of its huge imported Sanskrit vocabulary to express anything but the simplest ideas. This modern literary Bengali arose early in the 19th century, as a child of the revival of Sanskrit learning in Calcutta, under the influence of the college founded by the English in Fort William. Each decade it has become more and more the slave of Sanskrit. It has had some excellent writers, notably the late Bankim Candra, whose novels have received the honour of being translated into several languages, including English. Even he, however, sometimes laboured under the fetters imposed upon him by a strange vocabulary, and all competent European scholars are agreed that no work of first-class originality has much chance of arising in Bengal till some great genius purges the language of its pseudo-classical element.

Oriya Literature does not go back beyond the 16th century, though examples of the language are found in inscriptions of the 13th century. Nearly all the works are connected with the history of Krishna, and the translation of the Bhāgavata Purāṇa into Oriya in the first half of the 16th century still exercises great influence on the masses. Dīna Kr̥ṣṇa Dās (17th century) was the author of another popular work entitled Rasa Kallola, or “The Waves of Sentiment,” which deals with the early life of Krishna. Every verse in it begins with the letter k. It is not always decent, but is immensely popular. Upēndra Bhañja, Rājā of Gumsur, a petty hill state, is the most famous of Oriya poets, and was the most prolific. His work is insipid to a European taste, and when not unintelligible is often obscene. Oriya poetry, from first to last, has been an artificial production, the work of paṇḍits, who clung to the rules of Sanskrit rhetoric, and loaded their verses with so many ideas and words borrowed from that language that it is rarely understood, except by the learned. The whole literature is, in fact, overshadowed by the great temple of Jagannāth (a name of Krishna) at Puri in Orissa.

Assamese Literature.—The Assamese are justly proud of their national literature. It has an independent growth, and its strength lies in history, a branch of letters in which other Indian languages are almost entirely wanting. They have chronicles going back for the past 600 years, and a knowledge of their contents is a necessary part of the education of the upper classes of the country. In poetry, the Vaishnava reformer, Sankar Deb, who flourished some 450 years ago, was a voluminous writer. His best known work is a translation of the Bhāgavata Purāṇa. About the same time Ananta Kandali translated the Mahābhārata and the Rāmāyaṇa into his native tongue. Medicine was a science much studied, and there are translations of all the principal Sanskrit works on the subject. Forty or fifty dramatic works in the vernacular are known and are still acted. Some of them date back to the time of Śankar Dēb.

Authorities.—There is no work dealing with the three languages as a group. Both the Comparative Grammars of Beames and Hoernle (see Indo-Aryan Languages) are silent about Assamese. The fullest details concerning them all will be found in vol. v. of the Linguistic Survey of India, parts i. and ii. (Calcutta, 1903). In this each dialect and subdialect is treated with great minuteness and with copious examples.

The first Bengali grammar and dictionary in a European language was the Vocabulario em Idioma Bengalla e Portuguez of Manoel da Assumpçam (Lisbon, 1743). N. B. Halhed wrote the first Bengali grammar in the English language (Hooghly, 1778), but the real father of Bengali philology was the great missionary, William Carey (Grammar, Serampore, 1801; Dictionary, ib., 1825). W. Yates’s Grammar, as edited and improved by T. Wenger (Calcutta, 1847) and others, is still on sale. It is entirely confined to the literary Bengali of the paṇḍits. Its great rival has been Śyāmā Caraṇ Sarkār’s Grammar (Calcutta, 1850), of which there have been numerous reprints. In 1894 J. Beames published his Grammar (Oxford), now the standard work on the subject. It is largely based on Śyāmā Caraṇ’s work, but with much new material, especially that dealing with the colloquial side of the language. G. F. Nicholl’s Grammar (London, 1885) is an independent study of the language, in which the vernacular works of the best native grammarians have been freely utilized. There is no good Bengali dictionary. G. C. Haughton’s Dictionary (London, 1833) is perhaps still the best, but J. Mendies’ (Calcutta, about 1870) is also well known, and is the parent of countless others which have issued from the Calcutta presses. A Small Dictionary of Colloquial Bengali Words, by J. M. C. and G. A. C. (Calcutta, 1904), may also be studied with advantage. Cf. also Śyāmā-caraṇ Gāṇguli, Bengali Spoken and Written (Calcutta, 1906). For Bengali literature, see R. C. Dutt, The Literature of Bengal (Calcutta and London, 1895), and Hara Prasād Śāstrī, The Vernacular Literature of Bengal before the Introduction of English Education (Calcutta, n.d.). The most complete work is Bangabhāsā o Sāhitya by Dīnēś Candra Sēn (2nd ed., Calcutta, 1901) in the Bengali language.

For Oriya there are E. Hallam’s (Calcutta, 1874), T. Maltby’s (Calcutta, 1874) and J. Browne’s (London, 1882) Grammars. The last two are in the Roman character. They are all mere sketches of the language. Sutton’s (Cuttack, 1841) is still the only Dictionary which the present writer has found of any practical use. For Oriya literature, see App. IX. of Hunter’s Orissa (London, 1872), and Monmohan Chakravarti’s “Notes on the Language and Literature of Orissa” in the Journal of the Asiatic Society of Bengal, vol. lxvi. (1897), part i. pp. 317 ff., and vol. lxvii. (1898), part i. pp. 332 ff.

The first Assamese Grammar was Nathan Brown’s (Sibsagar, 1848, 3rd ed. 1893), and it is still the one usually studied. G. F. Nicholl gives an Assamese grammar as a supplement to his Bengali Grammar already quoted. Like that work, it is quite independent, and is not a revised edition of Brown. M. Bronson’s Dictionary (Sibsagar, 1867) was for long the only vocabulary available, and a very useful and practical work it was. It is now superseded by Hem Candra Baŗuā’s Hema-koṣa (Shillong, 1900). For Assamese literature, see Ananda Rām Dhekiāl Phukan’s A Few Remarks on the Assamese Language (Sibsagar, 1855), partly reprinted in the Indian Antiquary, vol. xxv. (1896), pp. 57 ff.  (G. A. Gr.) 


  1. In Mg. Pr. every r becomes l. For an explanation of the apparent non-observance of this rule in languages of the Eastern Group, see Bihari.