The History of the Bengali Language/Lecture 6

From Wikisource
Jump to navigation Jump to search



An analysis of the character of our vocalic and consonantal sounds can only lead us to form a definite idea regarding our provincial accent system. As such, a comparative study of Bengali phonology should be carefully pursued, though this sort of study involves a detailed consideration of the phonology of not only the ancient Vedic and the post-Vedic languages, but also of the Dravidian accent system. It is a hard work to do; but I must make an attempt in this direction, in the hope that my failure may lead to success, by attracting the attention and calling forth the energies of worthier scholars.

The letters of the ancient alphabet, which are not articulated in our pure Bengali speech, and which we keep in stock for the only purpose of representing the words of the ancient languages of India, will be dealt with only briefly.

অ.—(i) The usual sound of this vowel may be represented by 'o' as it is pronounced in pot, not, hot, etc. The sound of it in ancient times, may be presumed to have been like 'u' in but, cut, hut, etc., and so the long sound of this short অ may easily be made into আ (as 'a' in part, calm, large, etc.). This may explain why আ is the long form of অ, as ঈ and ঊ are the long forms of ই and উ. In Marathi, Canarese, Tamil and Telegu, অ is pronounced as short আ; this sound is exactly similar to our short unaccented আ sound as in আমার, আন্দাজ, আচার, etc. The Mundās and the Orāons of Bengal frontier do also pronounce অ as আ short. The Bengali sound of অ, as verging upon the sound of 'ও,' is naturally subdued when it is in the beginning or in the middle of a word, but it is distinct at the end of a word, when the final vowel, as a matter of course, is not হসন্ত in sound. Forgetting this natural mode of pronunciation, some writers unnecessarily write মতো for মত (like), ভালো for ভাল (good), কোনো for কোন (certain), etc. This process may be called wasting the black paint to paint a Negro black. The spelling reformers, moreover, fail to see that the sound of অ as a final, only approaches the sound of ও, and it will be creating confusion, if this sound be made identical with 'ও'.

(ii) The pronunciation of অ in Oriyā is midway between the Bengali and the Tamil pronunciation. The sound in Oriyā is similar to 'a' in ball; while we pronounce রবি as Robi, the Oriyās pronounce it as Raw-bi. This Oriyā pronunciation of অ is the highest exercise we make in producing the ancient sound, when reading Sanskrit, and this is the sound we produce, when accent is put upon অ not followed by ই or উ sound in a word. For example, when there is an accent on অ of অত (so much), the অ sounds like 'aw' in raw or like 'a' in ball. অ in অতি, however, does not change the normal Bengali sound, even though accented, as the ই sound follows it. Our spelling reformers may also take notice of the fact, that even when we attempt to maintain the purity of the sound of অ, as in অত, the sound is almost an 'ও' to the men of other provinces of India.

The Vedic sound for অ may no doubt be presumed to be our short আ, but we notice in the Atharvan Prātiśākhya (e.g. 1, 36) that besides the open sound, there was also a close or, saṁvṛta, sound of অ, which it appears, we have only inherited in Bengal. It is difficult to say, if this saṁvṛta sound has come down to us through Pāli and other later Māgadhi Prākṛtas, for we have no Prākṛta Prātiśākhyas to bear evidence to such a transmission. No doubt in Pāli, i.e., in the old Māgadhi Prākṛta, all nouns ending with the vowel sound of অ are found in the form নরো, ধম্মো, etc., in nominative singular, but this cannot perhaps be said to have been due to the peculiarity of the pronunciation of অ at the end of a word, for though there was no visarjania in use in Pāli, it may be said that, in its origin the sound came out of an elision of visarjania. It is, however, worth noting, that besides a general saṁvṛta sound for অ, we can detect in the Vedic itself a tendency of অ (as final) to be reduced to the sound of ও when joined to the visarjaniya: we first notice it very unmistakably in several euphonic combinations, where the final অ sound with the conjoined visarjaniya is reduced to ও; we again may notice that the dual form of দেবঃ, for example, is দেবৌ; the word দেবঃ must have been pronounced as দেবো (as in Pāli) for, to create a dual form by the lengthening of the final sound, the long sound of ও (which is ঔ) was reached, and this became the dual denoting suffix. It should be mentioned here, that the dual with ঔ is later in date in the Vedic language, and that the earlier বিভক্তি is noticed as আ in Chāndasa.

In consequence of their settled habit of pronouncing অ with its long and open sound, the people of Upper India, when pronouncing such words as 'long,' 'follow,' etc., by half adopting them in Hindi, utter those words as লাঙ্গ, ফালো, etc. The Bengali boys on the other hand, not being accustomed to emit the sound of 'i' and 'u' as in 'bird' and 'cut,' pronounce them as বা-র্ড (bard) and কা-ট, etc.

We gather from the works on Vedic phonology, that both অ and আ carried in their full-bodied open utterance, a half-distinct nasal sound. We can detect that the half-distinct nasal sound developed into a full-bodied nasal letter, when the sound in অ and আ was very long drawn to create, for instance, the plural form; this is why ভবতি became ভবন্তি, and we get in the Vedic speech জুহ্বন্তি as an additional plural form by the side of জুহ্বতি. How a দীর্ঘ sound naturally evolves a nasal, will be noticed presently. For want of any work on phonology, relating to the old Māgadhi speech, we cannot say, if the Prākṛta speakers of old Magadha displayed the peculiarity noticed in the case of the Vedic speech. It is, however, a fact that in the district of Bankura, as well as in the western portion of Burdwan, a final half-nasal occurs at the ends of such words as করিয়া, খাইয়া, etc., and they are pronounced as করিয়াঁ, খাইয়াঁ, etc. We also find that in some cases of our অপভ্রংশ words, where আ or any other vowel is accented to maintain the long sound, occurring in the original word, we put a half-nasal ঁ on the accented vowel, though the derivation of the word, does not justify the nasal. আঁটি (from S. অস্থি or অট্‌ঠি), জোঁক (from জলৌক), and সেঁতা (from সিক্ত) are some fitting examples. That in old Māgadhi Prākṛta a nasal was introduced in similar cases, can be presumed from some অপভ্রংশ forms; for example, from বক্র comes বঙ্ক, and from Vedic কুত্থ comes the form কন্থা; our vernacular forms বাঁকা and কাঁথা, I need hardly say, are from বঙ্ক and কন্থা respectively. We may note that বঙ্ক and কন্থা were adapted in Sanskrit from Prākṛta. সাঁপ from সর্প in Hindi may be compared with the above forms.

The nasal of আ.—I proceed now to show, that there is a natural basis in our very organ of speech, for the occurrence of this phenomenon, that আ carries at times a nasal sound with it. To do this, I have to also notice that আ is sometimes transformed into ই in the lengthening of the voice, as will also be pointed out presently. To serve our purpose, the results of a scientific research touching the origin and character of the vowel sounds, may be briefly stated here. Helmhotz and Koenig made very accurate and delicate experiments, in the mouth of men, and thereby accomplished a nice analysis of the natural sounds, produced by our vocal organs; it has been established by these experiments, that U (উ) is musically speaking the lowest, I (ই) the highest, and A (অ) the central of all the vowels. This scientific evidence in support of this proposition, that অ, ই, and উ are the three cardinal vowels, shows with what degree of accuracy, the old grammarians of India, analysed and classified the vowels, as well as the consonantal sounds, many centuries before the Christian era. We all know that অইঊণ্ is the first মাহেশ্বরী সূত্র, with which the old Sanskrit Grammar starts. It will not now be difficult to see, how আ becomes ই in the Vedic speech in the lengthening of that long vowel, some examples regarding which will be presently adduced. It has also been established by physical experiments, referred to above, that if অ or আ sound is lengthened without allowing the sound to reach a high pitch, that is to say, without allowing it to develop into ই sound, the lengthened voice is sure to become nasal; this is why আ carries with it a nasal when the sound is lengthened.

আ.—(i) Generally speaking our আ is a short vowel and it becomes long only when there is accent upon it. In respect of all long sounds, it has however to be noted, that there are different grades of them, and one sound, though long, may not be so long as another long sound may be. It is difficult to symbolise these grades of length, but we may formulate at least a 'half-long' sound, as intermediate between short and long. When আ is followed by a হসন্ত consonant, that is which does not carry a vowel sound, it becomes a half-long vowel as আ in আজ্, রাত, ভাত, পাত, etc. In the words আপন, কাপড়, পাতা, বারুই, বাড়ি, etc., আ is short; this short sound of a long vowel may be termed as the normal long sound in Bengali. We do not make any distinction between long and short sounds of vowels according to Sanskrit rules. As a rule, the single letters when uttered separately, as independent syllables or words, are uttered long; our children, unlike those of Upper India, pronounce অ—অ, আ—আ, ই—ই, ঈ—ঈ, ক—অ, খ—অ, etc., when learning the alphabet. Words of one letter are pronounced long, as it is the case in the Tamil speech. In Tamil, there are 42 one-letter words and they are sounded long. The Bengali one-letter word 'না,' to indicate a reply in the negative, is rather long, while it is short in না জানি, যাব না, etc., where it has been joined to, or compounded with other words. The initial letter of a word of two letters is pronounced long, when the final letter is হসন্ত; cf. ফল, শিব, গুড়, etc.; it is to be noted, that here the words of two letters are words of one syllable, and as such, the rule regarding one syllable is applicable.

(ii) I have spoken of the nasal sound which আ develops at times; it may be noticed that when আ carries a nasal sound in Bengali, it has the tendency to be pronounced long, no matter whether it is followed by a হসন্ত consonant or not; the long sounds of আঁ in আঁশটে, আঁচড়, কাঁঠাল, বাঁশী হাঁটু are examples.

(iii) In the Vedic speech, আ when made very long, was at times transformed into ই sound; compare সিধ্যতি and সাধতি (succeeds), শাস্তে (teaches) and শিষ্ট (taught); this is especially marked in final আ, as স্থিত from স্থা, গীত from গা, গীথ as a variant of গাথ, দীষ্ব from দা and হীন from হা (Macdonnell's Vedic Grammar, p. 4). We may notice in this connection, that in the অপভ্রংশ formation of আখ্ from ইক্ষু, the ই sound of the original has been reduced to আ in Bengali.

It will be noticed that in a large number of cases, the final আ followed by ই sound is changed into এ, as it is in the Vedic speech. Here the rule of law is, the preceding vowel sound influences the one coming after. This is the inverse of 'umlaut.' In our pronunciation মিঠা becomes মিঠে, রিষড়া becomes রিষড়ে, বিশ্বাস becomes বিশ্বেষ etc.[1] If however, the final আ comes after the উ sound, the আ is changed into ও, কুটা, পূরা, বুড়া, etc., are pronounced as কুটো, পূরো, বুড়ো, etc. If the final আ is attached to য় (which is pronounced as অ in Bengali) in a word of more than two letters, the whole of the final letter য়া is changed into এ or ও as the case may be, and this এ or ও takes the place of ই or উ of the next preceding letter; নদীয়া becomes নদে, বাঁশবেড়িয়া—বাঁশবেড়ে, করিয়া—করে, পড়ুয়া—পড়ো, লিলুয়া—লিলো, etc., গিয়া, দিয়া, ধুয়া, জুয়া, etc., being words of two letters they will be changed into গিয়ে, দিয়ে, ধুয়ো, জুয়ো (জুয়োচোর contracted into জোচ্চোর), etc. We thus see, that though 'য়' is seemingly pronounced as 'অ' the ই-অ sound is partly in our ears, to effect the phonetic change described above. Since the words are sure to be pronounced in the manner indicated above, even though they are spelt correctly in their original form, our spelling reformers may give up the attempt of spelling the words by representing all sorts of change of sounds, caused by phonetic decay. The writers, I speak of, want to introduce the verbs only in their contracted forms, but not the nouns, adjectives, etc. With what logic this distinction is sought to be made, is not easy to see. Why should we not write ইন্দিরে, ঈশেন, কৈলেশ, মিথ্যে, বিচের (as in আচায় বিচের), etc., if করে, গিয়েছে, etc., for করিয়া, গিয়াছে etc., may be introduced with propriety.

Having enunciated the rule regarding the change of final আ into এ, I must further note that the change spoken of, takes place even though other words ending with আ, and even when suffixes or inflections come after the final আ; মরেছিল from মরিয়াছিল, এল from আইল, বেঁধেছে from বাঁধিয়াছে, যেতে from যাইতে, etc., may be noted. No one should confound the forms গাইতে, নাইতে, চাইতে, etc., with যাইতে, পাইতে, খাইতে, etc.; in the former series, the letter 'হ' of the verb stems, has only lost the sharp aspirate sound, and as such গাহিতে, নাহিতে, and চাহিতে have been reduced to the softened down forms; it will therefore be ridiculous, if one would attempt a further contraction of the words of the first series into গেতে, নেতে, চেতে, on the analogy of যেতে, পেতে and খেতে, which are the contracted forms of যাইতে, পাইতে and খাইতে. We shall see from other examples later on, that even where we use the letter 'হ,' our tongue glides over the sound of 'হ' almost unawares, to allow the letter to perform its physiological work.

ই, ঈ.—I have said that there is no vowel in Bengali which has an inherent long sound; as such, ঈ and ঊ are used only to spell the words of Sanskrit origin. I have also noticed, that the vowels are sounded long, when we put accent or emphasis upon them, and they are all short when not accented. In pronouncing the Sanskrit word ঈশ্বর we do not really make ঈ long, but only a slight long sound is uttered as the result of our pronunciation of the compound letter which follows ঈ; all letters become slightly long in our pronunciation, when it is followed by a duplicated or a compound letter, no distinction can be made between the sounds of ই and ঈ when we pronounce ইচ্ছা and ঈশ্বর. Since we cannot indicate by phonetic representations, the long or rather accented sounds of অ, আ, এ, and ও, and since ঈ has not got a long sound in Bengali, it is hardly correct to write কী for কি to indicate accent or emphasis; for example, when we put emphasis upon the word তুমি (you) in such a sentence as আমাকে কে মারবে? তুমি? no one can seriously think of spelling the word as তূমি.

I have shown under the heading আ how this sound is changed into এ; it is also to be noted that when as a general rule, the aforesaid change takes place in our pronunciation, one exception is observed in some parts of Eastern Bengal. In the district of Dacca, the ই sound without being changed into এ, a metathysis in the sound occurs; for example করিয়া becomes কইরা, বালিয়া becomes বাইলা (বেলে = sandy), etc. It is to be further noted, that this phenomenon also occurs when the compounded ই-অ sound which is represented by what is called য-ফলা is taken by a consonant after the অ or the আ sound, and so মধ্য and বাধ্য are pronounced in some eastern districts, as মইধ্‌ধ and বাইধ্‌ধ. As a letter with য-ফলা is pronounced in Bengali as a duplicated letter, the ই sound comes before a compounded letter, specially when there is an aspirate sound at the end of the compound letter; thus the word ব্রাহ্ম will be pronounced ব্রাইহ্ম in Dacca. I must point out in this connection, that we meet with the forms কইল and ধইল for করিল and ধরিল, in our old Bengali books which were composed in the Rāḍha country. I have shown in a preceding section, that the Punḍra people, who inhabited the Rāḍha-land, proceeded to North Bengal in early times, and a large number of them migrated also to the Sambalpur tract, during the time of the Kośala Guptas. There may be or may not be any causal relation between this fact and what I am going to state now; it is curious, that unlike their neighbours on all sides, the people of Sambalpur reduce পানি, বালি, মালি, etc., to পাইন, বাইল, মাইল, etc.; a line of a song composed in the vulgar speech of the Rangpur district, will disclose the above peculiarity in that far off locality in Northern Bengal: আমার মইন্‌সা (মিন্‌সে—husband) ছিল ঘরতে (ঘরে) কি সতে (কি কারণে) আগ্ (রাগ) কইরা (করিয়া) গইছেন (গিয়াছেন) বাহে (মহাশয়) এক্‌না (একটা) কথাতে (কথায়).

We utter the ই sound in some cases to prepare the ground as it were, for pronouncing a compound letter, of which sibilant is a component part. The English word school is pronounced sa-kul in the Punjab, e-skool in the U. P., us-kul in some parts of Orissa and is-kul in Bengal, in the sea-board districts of Orissa, and in the Madras Presidency. I may remark in passing, that the disinclination to pronounce a compound letter as an initial, is India-wide; as the speakers of Aryan speeches in Europe pronounce the initial compound letters aright, and as it is a rule in the Dravidian speeches, that the initial letter can never be a compound letter, I am inclined to formulate a widespread Dravidian influence since a remote past, to explain this peculiarity in our pronuncia­tion. It will be noticed later on, that this inclination to drop the letter 's' as a first part of a compound initial letter, is noticeable in the Vedic speech as well. In the U. P., the introductory vowel sound becomes আ, when the initial compound letter terminates with আ sound, and so স্নান is uttered as আস্নান; in the case of other terminal vowel sounds, এ becomes the introductory sound. In the Punjab, the compound letter is split up, and one letter is pronounced after the other; in the Tamil pronunciation however, ই must be pronounced not only before the compound letters of the class spoken of above, but even before other initial compound letters; if even the second letter of a word is a compound letter, and the initial letter is a simple one, the ই sound is uttered by the Tamil people; the Tamil Apabhranśa of রাক্ষস is ই + রাক্‌কধন; this is also a rule in ihe Tamil speech, that an introducing vowel is added to a word beginning with র.

উ ঊ.—I need hardly repeat that ঊ has got no place on the list of pure and genuine Bengali vowels. উ is changed into ও, when as a pure vowel it is followed by আ or accented অ, in a word of two syllables; we may notice this change in ওঠ, ওঠা and ওড়া. In the Chittagong division, ও is changed into উ sound in some eases; I could not obtain a sufficiently large number of words to frame a generalised rule for this peculiarity of pronunciation in Chittagong.

In our vulgar Bengali pronunciation, a compound or duplicated letter with final অ sound, takes ই or উ when followed by ই or উ in the syllable, and so ইষ্ট, কৃষ্ট (কৃষ্ণ), মিষ্ট, etc., become ইষ্টি, কিষ্টি (in কেলে-কিষ্টি), and মিষ্টি, and পুষ্ট, তুষ্ট and উচ্চ become পুষ্টু, তুষ্টু and উঁচু; it will be noticed that at the dropping of one চ, the long sound of উ has generated the half nasal চন্দ্রবিন্দু. Consideration of the sound of ঋ will follow after taking notice of the sounds of য, র, ল and ব (v).

এ, ঐ, ও, ঔ.—These vowels or dipthongs have lost their original sounds and are pronounced as 'e' in met (or as 'a' in mate), as অই, as 'o' of the English vowel, and as অউ. The original sounds of these vowels have been partially retained in Tamil and Telegu; or rather, the Dravidian sounds of those letters nearly approach the ancient Aryan sounds of them. We learn from the re­marks of the ancient Grammarians (e.g., Vārtika of Kātyāyana, VIII-2, 106; under Mahābhāṣya, 1-1, 48), that এ is equal to অ + ই and ও is equal to অ + উ, and ঐ and ঔ being respectively the farther long sounds of এ and ও, the sound আ for অ has to be prefixed to ই and উ respec­tively, to obtain the proper sounds generated by ঐ and ঔ. The following examples collected by Prof. Macdonell in his Vedic grammar, may be profitably cited. We get in saptami singular অশ্ব + ই = অশ্বে, পদ + ঈ = পদে, and ভব + ঈঃ = ভবেঃ; notice also যমা + ঈ = যমে (twin sisters). When we get অনূপ (pond) from অনু + আপ, we see that উ has become long under the influence of succeeding আ. The two examples তস্ম + এ = তস্মৈ and দেব্যা + এ = দেব্যৈ are illu­minating. I may notice in passing, that in Orissa, ঐ is pronounced as এই and ঔ as আউ. It is clear that our Bengali pronunciation is wholly peculiar to us. ঐ and ঔ being merely long or augmented forms of এ and ও respectively, they ceased to be in use in the Prākṛta speeches.

এ.—(i) The sound of the vowel as indicated above is after the ancient Aryan sound of it, and this sound is uttered only in pronouncing the words which are at times designated by the technical term তৎসম. The initial sound of এ in indigenous Bengali words, can be represented by 'â' in mât. This normal sound of Bengali এ is so very exclusively peculiar to Bengali, that no letter or letter-signs of our ancient script, can represent it. It took me full six days to make a boy of the U. P. to pronounce the English word 'bat' correctly; the sound was altogether new to his ears and he was constantly varying his pronunciation from bet to bate. It is very important to note, that this â sound, as it occurs in bat, mat, etc., exists in the Dravidian tongue of Tamil; the pronunciation of the word কল্ (stone) as Kâil or of অন্‌বু (cajoling) as ânbu, are examples. Some Bengali writers, who do not know what the real sound of য় is, make at times the unscientific and useless attempt to represent the sound by য + আ to convey the peculiar Bengali sound to the ears of the foreigners. They do not know that the spelling গ্যাল for গেল is highly misleading to the people all over India, for ই অ is the sound of both য, and য ফলা—outside the presidency of Bengal. The Indians of all other provinces, as well as the Europeans, who come to learn Bengali after studying Sanskrit or Hindi, are sure to pronounce গ্যাল as গিয়াল. This method of representation will therefore be of no help to the outsiders, and will have the mischievous effect of teaching the Bengali boys a very wrong sound of য, which they have to deal with in their Sanskrit text-books. The Bengali boys learn the normal sound of Bengali এ even when they commence to lisp in their early infancy. All that we should do, is to frame rules regarding the normal and Sanskritic pronunciation of the vowel এ. As we have to learn the sounds of the vowels of the foreigners from the foreigners, our special Bengali sound has to be learned by those who are not Bengalis. To indicate this special sound of এ, we shall put a circumflex upon it.

I have stated that the initial এ in the তৎসম words, does not become the normal Bengali এ̂. I could notice only four words, namely, এ̂ক, ফে̂ন or ফে̂না বে̂লা and হে̂লা, which only seemingly form exception to the rule; we see that the old এক became এক্ক in the Prākṛta, and so the newly formed Bengali word এক, is not, inspite of its physical identity, identical with the original Sanskrit form; the words বে̂লা and হে̂লা are not of ছান্দস origin, and they were only adopted in Classical Sanskrit from Prākṛta; words with ফ initial are suspected to be non-Aryan in origin; ফাল, ফেন and ফল are the only words which occur in the Vedic language, of which ফেন alone belongs to the Aryan stock, being perhaps a decayed form of spena. I notice here that in the Northern and Eastern Bengal, the general tendency is to pronounce the initial এ of even the তৎসম words, as এ̂, and so কেবল, কেশব, দেশ, দেবতা, শেষ, etc., are pronounced with এ̂ initial.

এ as a single letter, and as an adjective, taking indicative particles as টি or টা, or particles of emphasis ই or ও, or being joined to other altogether separate words as রূপ, সকল, etc., retains its Sanskritic sound. The initial এ sound in a monosyllabic word, where the final consonant is হসন্ত, is pure; compare টের (to perceive) and টে̂রা (one who squints) and ঢের (much) and ঢে̂রা (clumsy as in ঢে̂রা সই). When however the final হসন্ত consonant is nasalised, the normal এ̂ sound will prevail, as কেঁক (a sound), খেঁক-শিয়াল, খেঁট (sumptuous feast), বেঙ্গ্, চেঙ্গ্ (a fish), টেঁক্, ঠেঙ্গ্, পেন্ পেন্ (slow teasing murmur), etc.

The words which are designated by some as তৎভব, including the অপভ্রংশ words, are subject to the rule governing the indigenous words. In Sir Rabindranath Tagore's highly suggestive work শব্দতত্ত্ব, the rule regarding the Sanskritic sound of এ, where normal Bengali sound might be expected, has been very nicely enunciated. His statement, that the এ sound when followed either by ই or উ sound, does not change into এ̂, is quite correct, the other rule formulated by him as noted below, should also receive universal acceptance. When the Prākṛta or অপভ্রংশ or দেশী root or stem of a word has an initial ই, the word derived from it with an initial এ will be pronounced with pure এ sound; কেনা begins with pure এ sound as the original stem or root is কিন্ (compare the forms কিনিলাম, কিনিয়াছিল, etc.), while বেচা having বেচ্ and not বিক্রয় for its origin, is pronounced otherwise; ঘেরা from ঘির্ (cf. ঘিরিল, ঘিরিলাম), চেরা from চির্ (cf. চিরিল), মেশা from মিশ্ (cf. মিশিতে) and মেলা from মিল্ (cf. মিলন, মিলিতে) are some examples. মেলা from মিল্ may be contrasted with মে̂লা (many) which claims no root having the ই initial. This will give us a clue to see that এমন, কেমন, খেলা, etc., are not the direct descendants of ইদম্, কিম্, ক্রীড়্, &c, but are derived from the Prākṛta forms এমৎ, কেমৎ and খেল respectively. It is to be noted that when 'ই' as a particle of emphasis, comes after the এ sound (এ-ই), the normal pronunciation is not changed; in এ, কি? however, এ is an independent word and so it retains the Sanskritic sound. Those who resorting to an unscientific method, do not keep ই and ও, indicating emphasis, distinct and separate from the words to which they are joined, must see on looking to the two foregoing forms, what a confusion may be created by their wrong spelling; if ই of emphasis is joined to হসন্ত-ক—final of the word এক in the shape of ই-কার, the two forms illustrated above will be identical. It becomes physically impossible to make ই or ও indicating emphasis, an inseparable part of such following words as তুমি, সে, আলো, etc., and one is bound to write তুমি-ই, তুমি-ও, সে-ই, সে-ও, আলো-ই, আলো-ও, etc.; why then such an exception should be made as to spell কোনো for কোনও when the emphasis itself is drowned in the new spelling, is difficult to understand. It must be pointed out, in this connection, that though য় is not pronounced as ই + অ, the sound ই is in our ears unconsciously, as it were since the general rule governs our pronunciation of such following words as কেয়া, খেয়াল, চেয়ার্ (chair), তেয়াগ, দেয়াল, ধেয়ান, পেয়ালা, শেয়ান্, হেঁয়ালি, etc.

Visarjaniya.—Neither the sign nor the sound of it was in use either in Pāli or in the later Prākṛta speeches; properly speaking, we do not use it in Bengali, and only three or four Sanskrit words as are spelt with visarjaniya, have been adopted in Bengali; these Sanskrit words again, are pronounced in the very fashion in which they were pronounced by the speakers of the ancient Prākṛta tongues; দুঃখ is pronounced as দুক্‌খ and নিঃশ্বাস as নিশ্-শাস্. Some of our Paṇḍits in their zeal and eagerness to give the Bengali language the sort of dignity which Æsop's jackdaw sought in the land of fables in dim backward of time, persuade us in vain to believe, that it is not মন but মনস্ which is the original word with us. They forget that we have borrowed মনোহর, মনস্তাপ, etc., in their entirety from what is called the Sanskrit language, and not that we have made euphonic combinations of মনস্ + হর etc., in our current speech, following the stringent Sandhi rule. The Sanskritists of the olden days borrowed মন্মথ ungrudg­ingly from a Prākṛta speech, and মনচোর is not found fault with, though চোর is a Sanskrit word; why should then our Paṇḍits shy at মন-কষ্ট or similar other forms in Bengali? Our natural sound has a history, and it cannot be obliterated. I may notice by the way (even though it is a digression) that in the Vedic speech we can notice a simple মন by the side of মনস্, as disclosed by such a term as আমনং signifying—having the mind or disposition to be friendly.

The nasal sound—The sound of the full-bodied final nasal is what may be represented by অনুস্বার; it is something like অ-ঙ্ in Bengali while it is অম্ in Sanskrit. In Telegu and Tamil there is an indigenous অম্ which is articulated as a final nasal at the end of all words, which are nouns. The Andhras and the Tamil people have not got the half-nasal ঁ in their script, and they do not also utter the sound involved by it; they have an abundance of 'am' and 'un' as terminal endings, but the half-nasal is altogether wanting. As the Mongolians are notorious for their excess of nasal sound, we cannot attribute the inability of the people of Eastern Bengal to utter the half-nasal sound to Mongolian influence. I may note that the half-nasal or ঁ is freely used by the people of Assam. The absence of the half-nasal must therefore be due to excessive Dravidian influence; in Eastern Bengal.

and ঞ—which are not strictly speaking independent letters, but indicate only the sort of nasal sound which must be generated, when occurring in conjunction with the letters of ক-বর্গ and চ-বর্গ respectively, have now lost their original sounds. সঙ্গ and বঙ্গ are pronounced almost as সং-অ and বং-অ; the full sound of গ is always subdued in গঙ্গা, রঙ্গ, সঙ্গ, etc. When ঞ sound comes first and the letters of চ-বর্গ are pronounced next, ঞ is pronounced as ন; অঞ্চল, বাঞ্ছা, অঞ্জন, ঝঞ্ঝা are pronounced as অন্‌চল, বান্‌ছা, অন্‌জন and ঝন্‌ঝা. When ঞ comes after জ, the sound becomes peculiar; জ which is no doubt a variant of গ, is replaced by গ, and the final nasal is uttered as half-nasal; আজ্-ঞা আজ্ঞা and প্রজ্-ঞা প্রজ্ঞা are pronounced as আগ্-গেঁ and প্রগ্-গাঁ in Bengal and Orissa.

ণ.—ণ is never sounded correctly in Bengal, and it is useless as a letter to spell the indigenous Bengali words. The use of ণ is limited within the sphere of spelling the Sanskrit words. It is worth noting, that though in the sea-board districts of Orissa, and in the eastern Garjat states, ণ is pronounced with considerable distinctness, the Oriyas in the Sambalpur tract, do not pronounce this letter, unless they are forced to produce the sound in schools, by abandoning their natural ন sound. Though we get enough of ণ in the script, representing Māgadhi speech of old, we notice the rule in the Prākṛta Grammar by Hem Chandra that ন takes the place of ণ in the Māgadhi Prākṛta. This is a very good evidence of the fact, that we have been pronouncing the cerebral ণ as ন since long. খোঁড়া (to dig) comes from the Prākṛta word খণণ, and hence a half-nasal and the cerebral ড় have evolved out of ণ sound. In কেষ্ট and বিষ্টু or বিষ্টুঁ for কৃষ্ণ and বিষ্ণু the cerebral ট has been substituted to represent the cerebral sound. The initial এ of কেষ্ট, comes from the tendency to pronounce এর্ for ঋ; it is no doubt a rare tendency, but it is noticeable in the pronunciation of ঘের্‌ত for ঘৃত, by some villagers, who attempt at Sanskritic pronunciation. I may notice, that the Iranian pronunciation of ঋ was এর্-অ in near approach of the old Vedic pronunciation of the vowel ঋ.

As we do not pronounce অন্তস্থ ব and identify it with ব of প-বর্গ, we do not keep the অনুস্বার separate from it, and compound it with ম which is the nasal sound of the letters of প-বর্গ; we write and pronounce বশম্বদ, কিম্বা and এবম্বিধ; this practice has been in vogue since long, as we meet with the above sort of spelling in ancient epigraphic records; I think this wrong use should be avoided, as it is necessary to know, which nasal sound is appropriate to which class of letters. We meet with such wroug spelling of words as তাম্ব and আম্ব in the old epigraphic records of East Bihar, Bengal and Orissa. The people of Orissa, pronounce আম্ব and not আম, and even now write তাম্ব্র for তাম্র in rural districts; it is noteworthy, that in some parts of Bengal আম is pronounced as আঁব and তামা as তাঁবা.

It is significant, that when ম becomes a ফলা or nasal adjunct of a consonant, it is not at all pronounced in Bengal, and the consonant with which it is compounded, is pronounced as if it is duplicated; পদ্ম and লক্ষ্মী are pronounced as পদ্‌দ and লক্‌খি, while our silent ম is distinctly pronounced in all other parts of India including Orissa. In some cases, it is noticed, that the consonant to which ম is conjoined, is dropped altogether, and ম is alone articulated; as from শ্মশান the word মশান has been derived. It is worth noting here, that this particular word "মশান" as the অপভ্রংশ of শ্মশান, is in use in some parts of Western India as well; the Pali apabhranśa form was "সুসান," from which our "শশান," has originated. I may notice one অপভ্রংশ of this class, which is in common use all throughout Northern India: শ্মশ্রু was first reduced to মশ্রু and then to মস্‌সু in old Prakṛta, and from this মস্‌সু the forms মোচ্ and মেছা came into use, and are still in use in some modern Vernaculars. I have shown in a previous lecture, that the peculiarity here noticed, is fully in accordance with the rule of grammar, as obtains in the Dravidian speeches; according to this rule, the initial letter of a word can never be a compound letter, and that double or triple consonants of different vargas, cannot occur anywhere in a word. The word জ্ঞান (জ-ঞান) has the pronunciation of 'gân'; but when the compound letter is a medial as in ধর্মজ্ঞান, the pronunciation will be 'Dhamma-ggân' without any nasal sound.

য, র, ল, ব.—Though these letters are regarded as consonants, their real character as compound vowels, has always been admitted by the old Sanskrit Grammarians. That য় is a compound sound of ই + অ has been pointed out before. As য is pronounced as জ in Bengali and Oriya, special mark has been added to the letter to signify the 'y' sound. It is worth noting, that in old Prakṛta speeches, we get জ for য and যৌবন for instance was pronounced as জোব্বন.[2]

র, ঋ, ল and ঌ.—The reason why I take up to discuss the sounds of the vowel ঋ and ঌ along with the discussion of the value of র and ল will be clear from the remarks I offer hereunder.

র.—I should explain that র originated from the compounding of ঋ + অ. It may be gathered from the Prātisākhyas (Rk. Prā-8-14; Atharvan. Prā-I. 37, 71) that the sound of a liquid was the final sound which ঋ produced. Again we get in the Vājasaneyī-Prātisākhya (IV. 145), that the initial half of ঋ had almost an অ sound. The Punjab frontier tribes pronounce ঋ as অর্-অ; the old Iranian pronunciation of ঋ, as may be detected in Āvesta literature, was অর্-এ and এর্-অ at times. We may notice that in ancient Prākṛta, বিকট became a variant of বিকৃত, কৃত became কত and মৃত became মত; this shows that ঋ had once the initial অ sound. The vulgar tendency to reduce ঘৃত to ঘের্‌ত and the pronunciation of কৃষ্ণ as কেষ্টো and of বৃষ as বের্‌ষো (as in বের্‌ষো-ছুগ্‌গু) remind us of the old Iranian pronunciation of ঋ. In the Rangpur district, the letters অ and র are interchangeable, when they are initials; this tendency is not wholly unknown among the populace in central Bengal as well. That the Prākṛta forms with অ and উ finals, were reduced to forms with র and রু finals may also be noticed, though in this case, the flat and boneless Prākṛta forms were really given stronger or more easily-pronouncible forms; thus we get মুনির, from মুনিঅ (gen.) and গরু, from গউ (গৌ). We may further notice, that when র is the adjunct to an initial letter of a তৎসম word, the sound এ is induced in pronunciation: প্রণাম, প্রসন্ন, প্রহ্লাদ, ব্রজ, etc., are reduced to পেন্নাম (পরণাম in H.), পেসন্ন (পরসন্ in H.), পেল্লাদ (পহ্‌লাদ in H.), বের্‌জ (বর্‌জ in H); contrast the forms মিত্তির, সুমুদ্দুর, বস্তর্, etc., from মিত্র, সমুদ্র, বস্ত্র, etc., where the letters with র-adjunct are not initials.

ঋ.—It becomes perfectly clear from the Prātisākhyas, that the Aryans in India were settled in their pronunciation of ঋ as 'ri.' Adverting however to such apabhramśa forms as বুট্টি from বৃষ্টি, উতু from ঋতু, etc., some scholars have wrongly asserted, that the Dravidian pronunciation 'Ru' for ঋ was in vogue in Northern India, when Pāli prevailed as a Northern Prākṛta speech. They have failed to see that only when the sound of ঋ had to be harmonised with the dominating উ sound in a word, that this vowel sound উ took the place of ঋ; we get ঘিয় or ঘি from ঘৃত where neither ই nor উ sound has to be assimilated; from ঋষি however, we get ইসি, while from ঋতু we get উতু because of the final উ sound. ব (v) is compounded of উ and অ; so the vowel ঋ conjoined to the accented ব (v) changes into উ; thus we get রুক্‌খ, বুড্‌ঢ or বুড়া, উষভ, etc., from বৃক্ষ, বৃদ্ধ, বৃষভ, etc. When however the final উ is not accented, and the letter joined with ঋ is accented, ঋ is not reduced to উ; for example, মৃত্যু has been reduced to মচ্চু, because of the accent on ঋ of মৃ preceding a compound letter.

ল—Grammatically considering ঌ generated ল, but this vowel never got any prominence. We should not forget to notice, that there was a field of a very free interchangeability of র and ল in the Vedic as well as in the post-Vedic speeches of the Aryans; this being an essential feature of the Dravidian speeches, the Dravidian influence in this matter as well is generally formulated. An additional 'ল' as a mixed sound of 'ল' + 'ড' occurs in all the Dravidian speeches; this 'ল' producing a cerebral sound, is in full use in Oriya and Mahrathi. Though this letter did not get admittance into the Vedic alphabet, the transformation of ল into 'ড' is recognised in the Vedic Grammar; we meet with the Vedic phonetic rule, that when 'ল' occurs between two vowel sounds, the letter may be optionally pronounced as 'ল' or 'ড' and so 'ঈ' লে may be uttered as 'ঈডে.' We have not got this cerebral in Bengali, but there are a few words in Bengali, which disclose the transformation of 'ল' into 'ড়'; তাড়ি (to-dy) the juice of Tāl (palm), may be an imported word, but কুঁড়ি from কলি (bud), মুড়া (extremity) from মূল, পাড়া from পল্লী (village), শিকড় (on account of the chainlike ramifications of the roots) from শিকল (—Sk. শৃঙ্খল, Pr. শিঙ্খাড় and শিঙ্খাটক) etc., are pure Bengali words. These forms, however, should not be confounded with those in which 'ড' or 'ড়' has originated from 'দ' or 'ত.'

Though ব (v) has lost its position in the Bengali alphabet, its উ-অ sound is retained in many words of va origin ; গুয়া from গুবাক, দেঅর from দেবর, দুয়ার from দ্বার, ধাওয়া from ধাব, জোয়ান from যুবন, and সোয়াদ from স্বাদ, are some in­stances. It mnst be familiar to the students of Prākṛita, that at times no distinction is made between 'b' and 'v' and that in the later Māgadhi, 'v' as an adjunct to a consonant, drops out altogether; such as, দ্বি is reduced to simple দি. The reduction of 'v' to 'b' cannot be said to be due to Dravidian influence, as distinction between 'b' and 'v' is strictly mantained in Telegu, Tamil, Malayalam and Canarese; but the dropping out of 'v,' when it is an adjunct to a consonant, can be explained by the rule of the Dravidian Grammar, which does not allow consonants of different classes to form a compound. According to this rule, দ্বি and স্বা of স্বাদ have to be reduced to দি or দু and to সো or সা, if they are initial letters; but if they occur as medials, they have to become দ্দি and স্‌সা respectively; compare the forms দুই, and সোআ, স্বাদ or সাদ on the one hand, and the pronunciation of the words অদ্দিতিয় (অদ্বিতীয়) and বিস্‌সাদ (বিস্বাদ) on the other.

Consonants.—As the question of Dravidian influence is being considered all along, I should add here a few remarks regarding the Dravidian alphabet system. In the Tamil script, we get only ক to represent all the consonants of ক-বর্গ and this ক is uttered with slight variations to pronounce খ, গ and ঘ. The usual Tamil sound of ক is almost খ to our ears. Similarly there are only ট, ত and প for all the letters of their বর্গ or class. There is only one letter to represent চ and শ and the sound of চ is peculiarly sibilant in all the Dravidian languages. In connection with the phonetic value of the Tamil consonants, a few remarks relating to the consonants of the Aryan speech, may be fittingly introduced.

From the admirable scientific analysis of the sounds of our letters, in old grammatical works, we get considerable information, regarding the genesis of the consonants. Professor Sayce, after considering the value of the grammatical works of the Greeks and other peoples, has rightly made this remark, regarding the Sanskrit grammar and phonology: "Far more thorough-going and scientific were the phonological labours and classification of the Hindu Prātisākhyas………The Hindus had carefully analysed the organs of speech, some centuries before the Christian era, and composed phonological treatises which may favourably be compared with those of our own day."

That ক changes into গ by slightly raising the accent (বাক্ + অর্থ = বাগর্থ) and that খ and ঘ are but aspirated sounds of ক and গ, need not be demonstrated; that চ is a variant of ক, ছ is an aspirated sound of চ, that জ and গ are always interchangeable, may be detected even on reference to the Sandhi rules. To serve the purpose I have in view, let me adduce here some examples from the Vedic or Chāndasa speech. From শুচ্ (glow of light) we get শুক্র and শুক্ল, both of which are identical in form and meaning, as র and ল are one and the same; we get also চিত্ (to perceive), কেত (desire) and চেত (= চিত্ত) lined together in one series. We may also notice, that from রোচ্ (bright with light) রোক (light) was derived; the later word আলোক owes its origin to রোক or লোক with an addition of আ as a prefix to the word. I draw the atten­tion of the readers to the words ভোজ (ভোগ), রুজ (রোগ), বিজ (বেগ) and ওজ, (উগ্র). Thus we see that the ক of Tāmil may fitly represent the whole series of letters of the ক-বর্গ; similarly ট, ত and প may be made as the sole legitimate fathers of the letters of their class. That শ is the same as চ may also be shown by phonetic analysis; in the old Iranian ছ, শ and স were interchangeable.

The pronunciation of the consonants of চ class is nearly as sibilant in Eastern Bengal, as it is in the Dravidian speeches; to represent this sound in letters I write here চ, ছ, জ and ঝ as—scha, ssa, dza, zha.

That the letters ছ, জ and ঝ were imported into ihe Tāmil script, some time after the introduction of what is called the 'Vatteluttu alphabet,' can be detected on comparing the modern Tāmil alphabet with the 'Grantha character' (prevailing now in Malabar) as well as with the Telegu script. The Telegu script, which agrees in the main with the Canarese, came into existence, at least as early as the 7th century A. D., since Hiuen Tsang speaks of the script, in the accounts of his travels in India. The epigraphic records show, that in the 6th century A.D., northern script was in use in the Tamilakam country, and that the modern Vatteluttu cannot be traced to a time earlier than the 8th century A. D. The Granth character was introduced in the 10th century A. D., to represent the North Indian Alphabet completely. At this day, the speech of Malabar was identical with the Tāmil speech, and some letters from the Grantha script were adapted in the Vatteluttu. These remarks will be sufficient to show, that when the Dravidian people first adopted the Northern script, they could, if they liked, introduce all the letters of the North Indian script; they elected to adopt a limited number of letters, so as to represent their natural sounds, they did not require all the letters for their use.

What I have stated in the previous lecture, of the origin and character of the cerebral letters, may just serve the purpose we have in view. I proceed now therefore, to consider the sound value of the sibilants and of the letter হ.

S.—It appears that ষ was derived from sibilant ছ which is intimately connected with খ in the matter of origin; লক্ষ্মী is pronounced লছ্‌মী in the U. P. and in Orissa, and খ represents ষ in many speeches in Western India; in the Canarese script ষ is written by giving one additional stroke to খ. Since ক্ষুদ্র, which is identical with ক্ষুদ্‌ল in the Vedic speech, became খুল্ল in early Sanskrit, to signify small (as in খুল্লতাত = খুড়া), we may safely assert, that ক্ষ was pronounced as খ in very early times, and this peculiarity is not due to corrupt pronunciation in Bengal. We must not forget, that in Pāli and in the later Prākṛtas, ক্ খ was written to represent ক্ষ as well as খ, occurring after a বিসর্জ্জনীয়. ষ is clearly pronounced as শ in Bengal and is never reduced to the sound of স. The sound of শ prevails in our speech, and this শ is pronounced as স, only in some cases where compounded with র and ল, as in শ্রী, অশ্লীল, etc. স is pronounced as a dental only when compounded with ত and থ. For purely Bengali words শ is the only sibilant that can be used.

Let me notice here some words of onomatopoetic origin, in which besides other sounds, ষ and ড played a good part. ষ indicated a heavy and solemn sound, while ড (ড় included) signified a roaring sound; বৃ and কং also conveyed or were made to convey a sonorous sound. বৃহ, বৃংহ, বৃংহতি signified the sound of a horse or an elephant; ঘোষ comes from ঘঃ (a heavy dull sound) + ষ; ঘণ্টা is composed of ঘং + টা (টা being the sound produced by striking something which is hard). Compare also the words মেষ ('মে' + ষ indicating sound), মেঘ, বৃষ, হ্রেষা, হর্ষ, ভাষ, মহিষ, রোষ (রু + ষ), কাংস্য (কং or ক্বণ্ sound + ষ), and শুষ্ক (from the sound শুষ্ occasioned at the drying-up of water on fire).

In the Sanskrit language of a comparatively later time, as well as in the Prākṛtas, ড may be pointed out as the letter which has been used to indicate an awe-inspiring sound; ষং + ড = ষণ্ড of the later Vedic speech, may also be noted. Our ঝড়, তোড়, দৌড়, কড় কড়, হুড় হুড়, মড় মড়, etc., are examples in point.

and শ.—The aspirate sound of হ which has created the extra consonants খ, ঘ, ছ, ঝ, etc., is not as is very distinctly marked, fully pronounced in Eastern Bengal; the subdued sound of it is something like অ—অ which cannot be properly pictured; হ as an initial letter, is very clearly pronounced in other parts of Bengal, but that there is a tendency to soften its sound when it occurs as a medial or a final, is to be duly noted. It is a peculiarity all over Bengal, that মহাশয় is pronounced almost like মশাই by dropping হ and by retaining a portion of the sound of the final য় or ই + অ; মহিষ is pronounced as মোষ by introducing the long ও sound compensating the loss of হ, and কহা appears in our speech and script as কওয়া. Even when we omit to write হ in such a word for example as তাহার, our tongue glides a little over the হ sound, and thus we can distinguish this word in our pronunciation from তার to signify either 'wire' or 'taste.' When হ takes a য-ফলা, it is pronounced as জ-ঝ in Bengali as well as in Oriya.

The pronunciation of ঘর or ভাত is not exactly গর or বাত in Eastern Bengal, but is something like গঅ-র or বাআ-ত, while in Dravidian pronunciation, they are uttered as গর্ and বাত্. The non-aspirated pronunciation of ঘ, ধ, ড, etc., seems therefore due to Dravidian influence. I remind you, that I noticed previously, the similar pronunciation of the people of Ceylon. In the Chittagong division of Eastern Bengal, however, the Mongolian influence has been so very much predominant, that in some points, regarding the articulation of sounds, the Dravidian peculiarities (though not obliterated) have been drowned; unlike their neighbours of the Dacca division, the people of the Chittagong division, breathe the 'h' sound into ক, চ, ট, ত and প, and pronounce ফাক্ (পাক্), খলম (কলম), ছুপ কর, (চুপ কর), etc.; not having done away with the original Dravidian influence, they do not pronounce ঘর and ভাত, but they pronounce them as গ্‌গর and ব্বাত by doubling the initial letters as it were.

Though the letters চ to ঝ are made markedly sibilants in Eastern Bengal, as they are done in the Dravidian lands, শ is seldom rightly pronounced by the ordinary people of Eastern Bengal; the reduction of শ to হ in Eastern Bengal (or more properly to a half হ with a wavy swing) cannot wholly be attributed to Mongolian influence, since such a change of sound, may be noticed in other parts of Bengal as well; the word গোশাল has been changed into গো-হাল or গোয়াল all throughout Bengal. In the Sambalpur tract, we hear হেইঠি (there) for সেইঠি of standard Oriya; this substitution of হ for শ is noticeable in Marhatti as well. I have already stated, that the sound of 'শ' predominates in Bengali; I should mention also, that Hem Chandra has noted in his Prākṛta Grammar, that 'শ' takes the place of 'স' all throughout, in the Māgadhi speech, though the representation in script of the Māgadhi Prākṛta, shows the use of dental স for the palatal শ.

Non-হসন্ত final.—We cannot conclude without noticing a peculiarity in the pronunciation of a simple consonant occurring as a terminal in a word. On reference to the list of words arranged in pairs below, it will be observed that final letters of the first words of the pairs are pronounced as distinct syllables, while those of the second words of the pairs are হসন্ত or silent. ভাল and ঝাল, ব্রত and মত (opinion), কাল (black) and লাল, ঘৃত and শীত, প্রমথ and শপথ, নিহিত and মোহিত, constitute the short list in question, to illustrate roughly this peculiarity. Let us frame tentative rules, regarding the pronunciation of the simple consonants, when they are final. We must first note, that as a rule, the final simple consonants are হসন্ত in Bengali, unlike what the case is in Oriya; the following are the rules for what form exceptions.

(1) When the penultimate is হসন্ত, no matter whether the penultimate and the final are made into one compound letter or not in spelling, the final is bound to be pronounced as a distinct syllable, unlike what is the case in Hindi; শক্ত, জব্‌দ, কষ্ট, etc., are examples. (2) The final simple consonants of the verbs in the Imperative mood, second person, are distinct syllables as in কর, বল, চল, etc., where the imperative-indicating final হ has now been dropped; when the expression is either non-honorific or highly honorific, হসন্ত sound prevails,—as কর্, বল্, চল্, etc. and করুন, বলুন, চলুন, etc. (3) (a) When the final letter is the representative of a compound letter of the original word, or (b) where the final letter of our vernacular word has become final by the decay of a syllable or of some syllables, স্বরান্ত pronunciation prevails. It is not asserted, that in all cases of such origin of words, the final simple consonant, must as a rule be pronounced as non-হসন্ত; what is pointed out is, that where the normal হসন্ত pronunciation is deviated from, the words disclose the history of their origin as formulated above. ভাল (from ভদ্র = ভদ্দ), এত (from এতাবৎ), বার (from দ্বাদশ = বাড়শ), etc., are some examples. Contrast খাল, ঝাল, এক, তিন, সাত, দশ, etc. Notice also মত (like) decayed form of our vernacular মতন and মত (opinion). It has also to be noticed, in the history of such words as ছোট, বড়, ভাল, etc., that their earlier forms were ছোটা, বড়া, ভালা, etc. কাল = black is pronounced as কালা in Eastern Bengal, and this form কালা obtains in Upper India. It is further noticeable, that আ, like ই-আ and উ-আ is found conjoined to many noun stems, to indicate the adjective forms of the nouns; we are not, however, concerned with that phenomenon here. (4) The participle-forming ত (but not ইত) is pronounced as a distinct syllable, as কৃত, ক্রীত, প্রীত, আগত, etc.; contrast with them রহিত, মোহিত, সহিত, etc. (5) The past-indicating ল suffix, which owes its origin to participle-forming ত, is pronounced non-হসন্ত, as করিল, গেল, হইল, etc.; the ছ­ending of the suffix আছ (occurring in second person only) is also similarly pronounced, as আছ, করিয়াছ, গিয়াছ, etc. (6) When the initial letter is compounded with র or ল, and (a) vowels other than অ, do not come between the initial and the final, and (b) the. consonant হ or ব does not intervene, the final letter is pronounced as a distinct syllable; e.g., ব্রজ, ব্রত, প্রমথ (contrast with শপথ of হসন্ত ending), শ্লথ, etc.; but notice the হসন্ত finals of (a) গ্রাস, গ্রাম, ঘ্রাণ, প্রাণ, প্রসাদ, প্রদীপ, শ্লোক, ক্লেশ, সু-গ্রীব, etc., where vowels other than অ intervene; mark again, (b) প্রহর and প্রবল, where হ and ব intervene. As an exception to the general rule, we get first, the word ধ্রুব, the final of which is uttered as a syllable; we notice the general exception, where ম is final; as ক্রম, ভ্রম, শ্রম, etc. (7) The simple finals of the words of two letters are স্বরান্ত, when the initial letter has ঋ for its adjunct; e.g., কৃশ, ঘৃত, নৃপ, মৃগ, etc.; contrast with them the হসন্ত sounds of the finals of কৃষক, কৃপণ, পৃথক, etc. (8) The finals of only a few re-duplicated words develop into স্বরান্ত sounds, when emphasis is put upon the words, as খচ-মচ in "কি খচ-মচই কর্‌ছ" and পড় পড় in ঘরখানি পড় পড় হইয়াছে.

In the name of framing rules, the cases where স্বরান্ত pronunciation prevails, have been set out in a classified order; to frame regular rules, we have to find out the essential underlying cause or causes, governing the phenomenon. In rule No. 1, we observe convenience in the matter of pronunciation. In rule No. 3, we notice, that to compensate the loss of letters at the end, a স্বরান্ত sound is drawn long; this is virtually the guiding cause in rule No. 2, since কর, চল, বল, etc., are the reduced forms of করহ, চলহ, বলহ, etc. As to other eases, I fail to enunciate any natural law, which causes the occurrence of the স্বরান্ত sound.

The scope of my subject did not allow me to notice the allied and cognate sound peculiarities as occur in other Aryan languages; I refer you however to a few Iranian peculiarities, just to suggest how wide our field of research is. In the first place we may notice, in connection with the sound of the vowel ঋ, that in Iranian, the radical 'ar' (ঋ) is reduplicated by ই; that the Indian sound of ঋ has always been 'ri' and not 'ru,' is clearly demonstrated by it. In the second place, we may observe without any reference to the ethnic composition in Iran, that 's,' is generally reduced to 'h,' which has been noticed as a special peculiarity in Eastern Bengal. In the third place, we may refer to the phenomena of Epenthesis and Prothesis, as occur in the old Iranian speech for comparison with similar phenomena, noticed before. As an example of Iranian Epenthesis, we may notice that the Vedic ভবতি stands as Bava-i-ti in Iranian; by Epenthesis, I mean the introduction of anticipatory ই or উ in the middle of a syllable. As to Prothesis, i.e., regarding the introduction of an anticipatory ই or উ, initially before a consonant, we may cite the example of ই ঋণ হতি which corresponds to Vedic ঋণক্তি. Many other Iranian peculiarities, as agreeing with some provincial peculiarities in India, may be studied very profitably by the Indian students of Comparative Philology.


  1. As in Eastern Bengal the last syllables are accented, the final আ of মিঠা and such other words, does not change into এ.
  2. In Bengali as well as in Oriya, য় as an initial or taking another য় as a ফলা is uttered as জ and in any other situation, it is য় as a rule. In Oriya, I could notice a single exception to the rule in the pronunciation of ময়ূর which is pronounced as মজুর. That this general rule, I have stated obtained in old Prākṛta, is pretty well known. It is noticeable in later Māgadhi speech, that the name জালি for example has been spelt with the usual জ initial, while this word occuring after উপ or rather compounded with উপ, উপয়ালি has been the form adopted.