The History of the Bengali Language/Lecture 7
SOME FACTORS RELATING TO PHONOLOGY AND ACCENT
A Comparative Study of Accent
The term akṣara (literally "undying," i.e., the ever-living and essential factor in human speech) signifies a letter as well as a syllable in the Vedic and so also in the later Sanskrit language. Different sounds of letters coalescing themselves in euphonic combination, and consonants unvitalized by vowel sounds, being joined to other consonants, generate compound letters; these compound letters, as well as the simple letters, being so many independent syllables in a word, must be separately pronounced. No doubt, in this method of pronunciation, we find the Vedic in agreement with the Sanskrit speech, but we have to notice that in the matter of accent, Vedic language differs very widely and radically from Sanskrit. In the Vedic language, the vowel sounds were not so very rigidly and unalterably fixed as long or short, as they are in Sanskrit; though, no doubt, a definite value is found assigned to each and every vowel, we can clearly see, on reference to the pada-pāṭha system, that the accentual stress of উদাত্ত, স্বরিৎ, and অনুদাত্ত, change what may be called the normal sounds of the vowels.
We have to first notice, that the final vowel of many flexional endings and of several adverbs, is given by the text, sometimes as short, sometimes as long. We have to notice next in the Vedic accent system, that not only the syllables, the words, and the phrases, but even many sentences are found accented. This fact, which discloses the living character of the speech, may be studied in the excellent analysis and discussion of the matter, in Prof. Macdonnel's Vedic Grammar. It may no doubt be said of the Vedic verses, that the general rythm of versification is not affected by accents; but that because of musical stress and accents, the verses are not lifeless quantitative ones (as in Sanskrit), should be duly appreciated. Since it is a fact, that natural gestures and modulations of voice, which contribute to the growth of the human speech, do survive as living factors in some proportion in each and every developed speech,—since it is undeniable, that every real and living speech must have an accent system of its own, a few examples of the Vedic accent should be adduced here, to form some notions regarding the Vedic, as well as the post-Vedic classical languages. Before citing the examples, I should note that উদাত্ত (as the meaning of it indicates) is the high accent in the Sāma Veda, while as a later innovation, স্বরিৎ of the next grade, is the high accent in the Ṛg veda; it will therefore be convenient to name the grades of accent by high, middle and low pitch or accent.
The first example I cite, is to show how by change of accent, a Vedic word changes its meaning. If the high accent be put upon রা of রাজপুত্র, the word will mean (being constructed as বহুব্রীহি) 'a person who is the father of sons who are or became kings'; but if the last letter ত্র is accented, the meaning will be, in the তৎপুরুষ compound, 'the son of a king.' For a similar change of meaning in Bengali, let me cite a few examples: if the Bengali word কলম is accented on the first syllable ক, the meaning will be 'a graft,' but with accent on the last syllable লম্, it will mean 'a pen.' How because of change in pronunciation, occasioned by difference in stress, a word varies in meaning, should be studied to realize the importance of accent in Bengali. Cf. আটা (flour) and আটা (gum), কড়ি (beam) and কড়ি (shell), কানা (blind) and কানা (edge), খোলা (open) and খোলা (tile), ছোঁড়া (boy) and ছোঁড়া (to throw), গেরো (a knot) and গেরো (as derived from গ্রহ signifying ill luck), চান (a bath, derived from স্নান) and চান্ (he wants), ঘাট (bathing place) and ঘাট (dereliction of duty), etc. Notice again, a case of accent where gesture becomes partly prominent. If a smell be pleasing, the word for our agreeable sensation will be normally accented, and the word গন্ধ will be accented on the first syllable, but our feeling of disgust about bad smell, will be expressed by putting a long accent upon the last letter, without any qualifying adjective being added: the utterance গন্ধ-অ-অ is sufficiently expressive. To express agreeableness, the particles of interjection in Bengali, are accented closely on the letter when the particle is of one letter, and on the first syllable, when the word is of more than one syllable; while on the other hand, in expressing our painful feeling or feeling of disgust, the accent is on the last syllable, and when the interjection is of one letter, the accent is put in such a manner on the letter, as to generate an additional syllable with a drawl sound upon it. For example, in expressing the feeling of admiration, the first syllables of বাহবা and বাপ্রে will be accented, and a close accent will be put upon the letter বা; on the other hand বাপ্রে (not as exclamation of admiration), is accented on the last syllable, to express the feeling of pain. Similarly আ! gets a broad accent, generating a drawl, to express pain or disgust. It may again be noted, that বাহবা! will be something like বা-হাও-বা, with accent on the second syllable, when there is a banter in the tone; to signify such a sentiment, the interjection বা will be so modulated, as to make it a word of three syllables, with two আ sounds. This is difficult to express in script. We may consider also, that the emphasis-indicating 'very much' tends to duplicate the final consonant of a word; অত্ত from অত and এত্ত (so much!), এট্টু (very small), কোত্থাও, গোচ্ছা from গোছা, ছোট্ট, ঝক্কি from ঝুঁকি or ঝোঁক, ধাক্কা and ধক্কল, ফক্কা from ফাঁকা, বড্ড, বক্কেশ্বর from বকা, রগ্গা from রোগা, রত্তি, as in একরত্তি, from রতি, and সক্কাল বেলা (very early in the morning) are examples. We may compare similar forms in Oriya of Sambalpur, as খড্ডা (to fry) to indicate বেশী খড় খড়ে করে ভাজা (to make crisp by overdoing), and মেচ্ছা (for মেছা = moustache) মুরুছি (twirling) to express one's defiant attitude.
For my second example, regarding Vedic accent in metrical composition, I quote a Vedic verse, which is full of emotional sentiments. In this verse, the mixed feeling of eager solicitude and despondency has been expressed. The first portion of the first verse of the 95th Sukta of the 10th Mandal, which is addressed by Pururavā, to his fugitive wife Urvaśi, on meeting her accidentally, stands with accents as follows:—
"হয়ে জায়ে মনসা তিষ্ঠ ঘোরে
বচাংশি মিশ্রা কৃণবাবহৈনু"
We cannot fail to notice, that many long vowels have been made either short or semi-short, with accents of lower grade upon them, and the short syllable ন in মনসা has a high accent on it. How on account of the subdued utterance of 'হয়ে,' and a high accent on the final syllable of 'জায়ে,' the feeling of coaxing with fervour, has been expressed, may be explained and appreciated, if in the first place, the verse is correctly recited, and in the second place, we carefully consider, how to express this very sentiment, we accentuate our words and modulate our voice to-day. If we translate হয়ে জায়ে by ওগো উর্ব্বশী or by ওগোও, we can see that to express anxious solicitude or cajoling, we have only to half accentuate ওগো, but have to fully accentuate শী of উর্ব্বশী, or the last ও of ওগোও, with a peculiar modulation of voice. This thorough agreement of Bengali accent, with the Vedic, in this particular instance, may be merely a chance agreement, but all the same, it is interesting to note, that in Hindi, as well as in Oriya, the method of accent is different. In the corresponding Hindi form of the expression, as য়ে উর্ব্বশী or ইয়ে পিয়ারে, the interjection ইয়ে will be pronounced with high accent or great emphasis; similarly 'এ' of এ উর্ব্বশী or এ ধন of Oriya idiom, will require the high accent to be placed on এ. Even though ও comes before a name in the vocative case in Bengali, the name itself is modulated peculiarly, to signify address, and the interjection portion is not so vigorously pronounced, as it is done in Hindi; the forms, রাম হো (Hindi) and হৈ হো রাম (Sambalpur Oriya) may be compared with our corresponding Bengali form. In the Nepalese, এ must invariably come before a word in the vocative case. When thus noticing different accentual peculiarities in the vocative case, I should note, that in the Dravidian speeches, interjections do not occur before the words in the vocative case.
The third example, I cite for Vedic accent, relates to the pronunciation of the word Agni, as occurs in the Sāma Veda, the Maitrāyani Samhitā, and the Kāṭhaka Samhitā; we find the high accent placed upon the compound letter গ্নি (i.e., to say upon the final নি) and not upon the initial অ, which alone should be pronounced long in the Sanskrit language. That with this very accent on the second syllable, the word অগ্নি was pronounced in our oldest-known prākṛta (misnamed Pali), may be gathered from some facts which I should notice here. We get গিনি and অগ্গি, as two different decayed forms of অগ্নি in the aforesaid Prākṛta; they represent presumably, two different provincial forms of the old time. We can clearly see in the history of the word গিনি, that on account of accent on the last syllable, the unaccented first syllable dropped out, following the natural rule of phonetic decay. In our consideration of the form অগ্গি, we first notice, that according to our previously formulated rule (stated in illustrating the Dravidian influence), গ of the second syllable has been doubled, as letters of different বর্গ cannot form a compound; the word is therefore, more in a changed form than in a decayed state. As to the pronunciation of it, I refer to a line of a verse composed in ইন্দ্রবজ্রা, as occurs in the Thera Gāthā: —অগ্গি যথা পজ্জলিতো নিসীথে (like the fire blazing in the midnight). The metre here is only seemingly faulty, as the second syllable of অগ্গি is not long; but if we accept the proposition, that the short syllable in question has a high accent on it, it will be admitted that the accent makes up for the shortness in question.
We notice, both in the so-called Pāli and the later Prākṛta dialects, that their speakers changed the spelling of the words according to accent and pronunciation, but did not introduce like the Dravidians, such additional vowels as short এ and short ও. The fact however, that long and short vowels were articulated short or long, following the natural accent, can be easily determined by referring to the prose composition with which the Prākṛta literature abounds. Professor Pischell has rightly asserted in his work on the Prākṛta Grammar, that the Vedic accent or tone did not die out, but existed in the so-called Pāli Prākṛta. The learned scholar came to this conclusion, by looking deep into the causes, that led both to the phonetic decay and the accretion of new letters in the Pāli words. Professor Jacobi, misled by the modern artificial method of reading Pāli, has criticized this view, and has asserted, that the Vedic tone did not survive in the Pāli speech, but only it was a sort of stress, that was in use in the utterance of words. We must remember, that accent is a thing of hard growth and cannot die out easily. By the introduction of new racial elements, the old accent system of a speech may undergo some change, but the system itself is not wholly effaced.
We have seen, that a very regular and thorough-going accent system prevailed in the Vedic speech; we have also seen from an example of a Vedic verse, that emotions were freely expressed in the Chāndasa speech, and in consequence thereof, the vowels could not be kept rigidly fixed in their long or short pronunciation. It is on the other hand perfectly clear, that the Vedic text for illustration (beginning with হয়ে জায়ে, etc.) will be a lifeless quantitative verse in Sanskrit, requiring artificial and meaningless raising and lowering of vowel sounds in the following form:
˘ ¯, ¯ ¯, ˘ ˘ ¯, ¯ ˘ ¯ ¯, etc.
We should not however overlook, that inspite of rigidity of rules, we have to put different stress upon different words occurring in a Sanskrit verse, when there is a feelingly recitation in contradistinction with what may be called metrical articulation. No doubt, we can see, that this sort of putting emphasis upon phrases, is not due to the living character of the language, in which the poems are composed; there are many good Sanskrit verses, wherein we find, that their poets by virtue of their skill, have arranged the words of long vowels in such a manner, that the feelingly expressions may be appropriately brought out, by putting emphasis upon the long vowels only: The example of the verse in the নীতিশতক, beginning with নিন্দন্তু নীতি নিপুণা যদিবা স্তবন্তু, is in point. This is rather infusing life in dead bones.
I have judged here the Sanskrit language, by taking the question of accent only into consideration; other facts, as are necessary to be discussed in determining the character of a language, will certainly be discussed relevantly in a subsequent lecture, but we should not forget, that accent is the life breath which vitalizes the words, and a speech without accent is a contradiction in terms. We have seen from the accent point of view only, how unnaturally rigid the Sanskrit language is; we will see on references to other facts, that this artificial rigidity is due to the fact, that Sanskrit had to accommodate itself within a frame-work of generalized rules, which some mighty grammarians constructed in their zeal to perpetuate linguistic purity, when in consequence of a natural change, the speech of the holy Vedas was transformed into a new popular speech.
I should mention in this connection, that some persons are very wrong in their opinion, that such a living speech as Oriya, is without an accent system. The misconception is due to the fact, that in Oriya, as in Sanskrit, all the letters are pronounced as distinct independent syllables. That inspite of it, the letters and words are accented by the Oriyas, without any reference to the long or short sound of a vowel, is instructive. I cannot deal with the Oriya accent system here; I adduce only one example to serve my purpose. গলা (gone) as an exclamation of surprise, will be articulated with accent on the first syllable, which is short, and the word will sound like গ-অ-অলা (Oh! it is gone), while the accent on the second syllable in similar feeling of surprise will bring in another গলা as প্রচয় or accumulated unaccented letters, like a tag to the word, and the sound will be like গলা-আগলা. I should inform you in this connection, that in Oriya the final syllable of a word is generally accented and this is why, (a) the final simple consonants are not pronounced হসন্ত, (b) রাজা has been reduced to রজা and the Apabhraṁśa form of কাংস্য is কঁসা, and (c) the half nasal ঁ is placed on the final letters of such words as কাহিঁ, যহিঁ, etc.; it is noticeable, that when the Bengalis write the Oriya words কাহিঁ, যহিঁ, etc., they represent them as কাঁহি, যঁহি, etc.
In my general criticism relating to the question of accent, I have noticed many peculiarities of ours; I proceed now to consider some other marked peculiarities, as should engage the attention of all scholars. As a general rule, in our standard Bengali speech, the first syllables are accented, while curiously enough the last syllables are accented in Eastern Bengal. Mr. J. D. Anderson, late of the Bengal Civil Service, has written of late some very suggestive and learned notes in the J. R. A. S. on the character of Bengali syllable and accent. To the students of philology these original notes are of very high value. Misled by the wrong idea or information, that the people of the districts of Mymensingh and Dacca, have the tendency to make the first syllable accented, Mr. Anderson has compared the Assamese mode of pronunciation with that of Eastern Bengal, and has subjected the Eastern Bengal accent system to Assamese influence. As the case is quite the contrary, we must look to some other influence for this phenomenon. Let me just give a few examples as to how the words are accented differently in Eastern Bengal and Central Bengal. In Central Bengal, মাথা having the accent on the first syllable, the unaccented syllable is pronounced soft, and থ sounds like ত; in East Bengal, however, মার্থা, কর্থা, কের্রে, etc., are the accented forms. With accent on the first syllable, the final য় of কোথায় is wholly or partially dropped in Central Bengal, while the accent on the last syllable in Eastern Bengal brings out থা and য় distinctly. As a result of accent being placed on the first syllables, the unaccented second syllables of many words, have undergone a natural phonetic decay, in the common speech of the people of Central Bengal, and thus the unaccented ই sound in the second syllables of করিব, করিয়া and করিলাম have become exiinct, and the forms করব, করে and করলাম have prevailed. As in Eastern Bengal the last syllables of আইল, আইস, etc., are accented, and the first portions must be uttered to come to the last syllables, almost no change takes place in those words; but when the first portions are accented, the two vowels আ and ই, coalesce and long এ, which is the combination of আ and ই takes the place of the first two letters; thus এল and এস take the place of আইল and আইস. In Manbhum and in some parts of Bankura, which are contiguous to the lands of the Dravidians, the last syllables are mostly accented; in the sentence তোমার বেটা বটেক্? (Is he your son?) the last syllables of all the three words are accented. Because of change of accents, বেটা is pronounced বেটা in Central Bengal. কুত্থ, the old Prākṛta or proto-Bengali form of কোথা still survives in Manbhum, partly because of the hilly accent, and partly because the accent is placed on the last syllable. It is notorious, that the last syllables of words are generally very much accented by the Dravidians. The final আ of foreign nouns are for this reason made into āi, as for example মণিমেখলা becomes মণিমেখলৈ. This is why the vulgar people in the South, pronounce the English words 'government,' 'and,' etc., as governmen-ta, an-d, as, etc. I have spoken before of the Dravidian element in Bengal; it is the excess of this element which I suppose to be the cause of Eastern Bengal peculiarity. In the Chittagong Division however, where the Mongolian influence is considerable, the first syllables are mostly accented vigorously, by almost duplicating the accented syllables, and thus পাক, তোমার, etc., are reduced to ফাক, থোমার, etc.; but in other respects the widespread Dravidian influence of basic character, peeps through the thin Mongolian veil.
I have spoken before of the genera] disinclination in Bengal, of not fully articulating হ, when it is not an initial letter, and that thus we have got such forms as গাইতে, সইতে, রইতে, etc., for গাহিতে, সহিতে, রহিতে, etc. When the tendency to put the stress of accent upon the first syllable is coupled with this phenomenon, we see how তাহার and তাহাতে are reduced to তার and তাতে. That in spite of the decay of হ, the sound of the letter is partially retained in our tongue, can be detected in the mode of our pronunciation of those words, in contrast with the pronunciation of তার meaning taste or wire and তাতে, a verb which signifies "becomes hot." Really speaking therefore, হ is not dropped, but its sound fades into indistinctness, after the accented তার; as such, it is improper and useless to leave out হ in our spelling in these cases; those who elect to pronounce তার for তাহার will do so in spite of the letter হ, for accent on the first syllable will soften the sound of the letter হ. The so-called reformers should see, that if হ is retained, it will not be sounded much because of the accent on the preceding syllable; again presence of হ will maintain the right pronunciation of the words concerned.