4.0. Introduction
Hitherto, the main concern of almost all the studies of Bangla metrics is restricted
to Bangla poetry. Since at least 1914 the metrical structure of Bangla poetry
has drawn the attention of a fairly good number of metricists - a remarkable event
that inevitably enriched, on the one hand, the entire field of Bangla metrical
practice by giving rise to a wide range of varieties therein under two main types,
viz. conventional and experimental, and on the other hand, the literature on Bangla
metrical theory by adding a good number of analyses of Bangla metrics done from
various viewpoints.
The metricists in Bangla may be divided into more or less two categories:
a)
the poet metricists, and
b)
the grammarian metricists.
In
the first category mention may be made of Tagore, Dwijendralal Roy, Satyendranath
Dutta, Sasanka Mohan Sen, Vijoy Chandra Majumdar, Dilip Kumar Roy and so on -
all of whom are creative poets of Bangla. Even as metricists they dealt with Bangla
metre more from the view point of a creator rather than that of a pure grammarian.
According to Prabodh
Chandra Sen, henceforth Sen, (1986: 23):
"from
the essays by Tagore, Dwijendralal, Vijoy Chandra, Satyendranath and Sasanka Mohan
I gathered the impression that it may not be impossible, but certainly not be
easy for a born poet to differentiate between the art and the grammar of a poetry.
In other words, the viewpoint of a sheer grammarian is neither expected from a
born poet, nor should it be. Hence non-poets ought to come forward to perform
this job."
In order
to fill up this gap, i.e. the grammarian's view point, a second type of metrical
analysis came into existence. In this second category mention could be made of
Sen and Amulyadhan Mukhopadhyay, henceforth Mukhopadhyay.
Sen
(1986: 23), in his introduction to metrical analysis, says: "I, in spite
of great hesitation, totally depended on my own grammarian's intuition and ventured
upon such an analysis."
Even
about Mukhopadhyay, Sen (1986: 309) admits that, "Amulyadhan, however, enriched
the metrical literature as a pure grammarian metricist."
To
summarize, it may be said that the essays by the metricists belonging to the first
category, i.e. the poet metricists, represent the creative viewpoint of an artist;
whereas those by the second category, i.e. the grammarian metricists, represent
the analytical viewpoint of a grammarian. For the present analytical purpose I
need to consider the works by the second category that deal with Bangla metrics
in terms of a grammarian's analytic viewpoint.
Among
the works of Mukhopadhyay I shall consider his baNla chOnder mulSuttro (1957)
as the most representative one as it is a non-controversial item included in the
undergraduate syllabus for a fairly long time.
In
the massive corpus of research by Sen, stretching from 1922 to 1986, according
to his own judgement nutOn chOndo porikkrOma (1986) is the most mature statement.
In his own words (1986: 30) "It won't be an exaggeration to say that this
book (i.e. Sen, 1986) is the most mature outcome of the metrical ideas and practices
of my whole life".
Between
these two works, though Mukhopadhyay (1957) is pedagogically more appealing than
the other one, I shall report the indigenous work on Bangla metrics solely on
the basis of Sen (1986), i.e. the Sen school, because of the following reasons:
1.
Sen's metrical analysis consists of three autonomous levels, viz. phonological
nature of Bangla metrics, rhymes and its types, and morphology of Bangla metrics.
In his own words (1986: 28) "each chapter is autonomous…three chapters
represent three hierarchical levels".
Apart
from this, in the first chapter Sen deals with the basic concepts such as pause,
stress, syllable, quantity-unit, mora etc., of the theory very objectively. Such
objective notions and explanations are the cardinal elements of the current field
of metrical phonology.
On the contrary, Mukhopadhyay (1957) does not introduce
any such autonomous levels or basic concepts in his theory - as a consequence
of which his theory fails to contribute directly to the field of metrical phonology.
For example, in this theory the syllable and the quantity-unit are treated as
two metrical-pattern-dependent concepts, rather than two independent basic concepts.
2. Compared to Mukhopadhyay
(1957) Sen's theory is more ubiquitous in application and free from self-contradiction.
Mukhopadhyay, with respect
to the method determining the quantity-unit for the syllable, divides syllables
into two broad categories, viz. normal syllables, which exhibit normal quantity,
or agree with the normal pronunciation of Bangla, generally available in prose;
and deviant syllables, which are the residual or non-normal types not available
in prose.
This approach shows that Mukhopadhyay's concept of syllable
applies only to the field of Bangla poetry; whereas that of Sen applies to the
field of the Bangla language as a whole.
Moreover, Mukhopadhyay (1957: 57) says that "keeping the pronunciation-style
intact, the quantitative unit of the syllable is decided in accordance with the
metrical patterns".
These two are contradictory statements because in terms of the earlier statement
only the normal quantitative unit is determined by the normal pronunciation style,
not the deviant one. But the metrical pattern determines both.
Sen (1986), however, is free from such self-contradiction.
3.
Over a period of little more than sixty years (1922 - 1986) Sen developed a scientific
and transparent theory of Bangla metrics through constructive criticisms and revisions.
And in this process even a school, which may be termed as the Sen school of Bangla
metrics, has come into existence. Among many other metricists few well-known names
in the Sen school are Ram Bahal
Tewari, Shajahan Thakur, Hassan Al Abdullah,
and Pabitra Sarkar.
Though
because of the repeated revisions the Sen school is not very popular with students,
i.e. the pedagogical approach of the Sen school is not quite effective, the researchers
often feel comfortable with the analytical style of this school.
On
account of the above reasons I prefer to focus on the Sen school in the present
chapter.
It has already
been mentioned that Sen (1986) discussed Bangla metrics in three different autonomous
levels in three chapters. The three chapters are: (1) the phonological nature
of Bangla metrics, (2) rhymes and its types, and (3) the quadruplicate format
of Bangla metrics. Among these three Sen terms the first and second jointly as
the "phonology of Bengali versification" and the third as the "morphology
of Bengali versification".
In
his own words (1986: 28): "…three chapters represent three hierarchical
levels…the first chapter may be considered as the introduction to the other
two. In fact, this chapter is the synopsis of the overall metrical grammar".
In
accordance with Sen's own judgement, in the present chapter I shall discuss the
basic concepts of the first chapter of Sen (1986).
Section
1 will discuss pause and accent; section 2 will deal with the syllable; section
3 will illustrate the metrical styles, which will be followed by the concluding
section.
4.1. Pause and
accent
Well-restrained
and well-measured arrangement of sound is rhythm. If the speech sounds of our
normal prose are regulated and arranged in a well-measured manner we get the rhythm
of a verse.
The form
of the prose language is regulated mainly in terms of motion and pause. A portion
of prose limited by a motion-pause couple is called the speech-foot. Each speech-foot
starts with a motion and ends in a pause. And the auditory undulations of speech
caused by this pattern of motion and pause are termed as rhythm.
Both of the modes of stylistic organization, viz. prose and verse, consist of
rhythmic sequences. The prose rhythm is undetermined in nature as the prose feet
are unmeasured; while the verse rhythm is well-determined in nature as the verse
feet are well-measured. The primary concern of metrical literature is to determine
the principle of measuring a foot, by analyzing the foot-formation of the verse.
The foot contains a
motion that starts with an accent and ends in a pause. Accent is of two types,
viz. pitch accent and stress accent. Since Bangla contains only stress accent
hence here the term accent will mean stress accent only.
Accent and pause may be exemplified as follows:
1.
|cOndro jOkhon | |Oste namilo || |tOkhono roYeche | |rati,
|purbo diker | |OlOS nOYone || |meliche rOkto | |bhati
(Sindhupare/ cittra/ Tagore)1
"moon
when horizon-loc. got-down then-incl. is-continuing night
eastern side-gen.
idle eye-loc. is-opening red radiance"
'Night
was still there when the moon was about to set below the
horizon and the
idle eyes of the eastern side started glowing red'.
Here
accent and pause are indicated by the signs [ | ] and [ |, || ] respectively.
In
the verse rhythm of Bangla, pause plays a more important role than stress. Sen
supports this view. In Sen (1986), firstly, pause has been dealt with in an elaborate
manner, and secondly, the different segmentations of the rhythmic verse too are
pause-dependent.
Even
Mukhopadhyay (1957: 218), in support of this view, says:
"In
English pronunciation accent is the most important prosodic factor so the nature
of rhythm here depends on the placement of accent. As in English prose-rhythm,
so also in English verse-rhythm accent is the most prevalent phonological cue.
But in Bangla the nature of rhythm depends on the placement of pause. Bangla rhythm
is identified in terms of quantity, measured in terms of feet, between two pauses.
The process applies to both types of rhythm, viz. prose-rhythm and verse-rhythm."
As
pause plays a more important role than accent in Bangla rhythm, here it will be
discussed before accent.
4.1.1.
Pause (joti)
In this
section I shall discuss pause types; metrical units in accordance with different
types of pause; location of pause; and elision of pause.
4.1.1.1.
Pause types
In terms
of a hierarchical arrangement, in poetry pause may be of five types. From the
weakest to the strongest these five types are grade 1/ dOlojoti, grade 2/ upojoti,
grade 3/ loghujoti, grade 4/ Ordhojoti, and grade 5/ purnojoti.
Each line of the example no. 1 contains 4 pauses belonging to different grades.
For example, though the first and the third pauses, indicated by the signs [|],
are quite distinct, they are weaker than the second and the fourth ones. To be
more specific, the second pause, indicated by the sign [||], is apparently stronger
than the first and the third ones. And the last, i.e. the fourth one, not indicated
by any special sign, is the strongest of them all.
In
the metrical terminology the first and the third pauses represent grade 3; the
fourth pause represents grade 5; and the second pause represents grade 4, i.e.
it stands between the other two.
More
careful listening, coupled with slow (lento) tempo of enunciation, reveals that
even within the foot there is a pause, somewhat weaker than the others. Such intra-foot
pause represents grade 2 and is indicated by the sign [:], e.g.
2.
cOndro : jOkhon | Oste : namilo || tOkhono : roYeche || rati,
purbo : diker | OlOS : nOYone || meliche : rOkto || bhati
(cf. no. 1)
Moreover,
a consideration of the minute details of pronunciation will show that in actual
speech even the words are broken into small pieces, for example, some words contain
a single piece, such as na 'not', ki 'what', Se 'he', maS 'month', phul 'flower'
etc; some contain two or more pieces, such as no.di 'river', a.kaS 'sky', ko.bi.ta
'poetry', bO.Son.to 'spring' etc. Each such piece is followed by a minute pause.
Such a minute pause, indicated by the sign [.], is the weakest type in the pause
hierarchy and thus represents grade 1; these 'pieces' of word, in metrical terminology,
are called the syllables. These five types of pause - grade 1 marked by [.], grade
2 marked by [:], grade 3 marked by [|], grade 4 marked by [||], and grade 5 indicated
by the absence of any sign - are ordered hierarchically. If we use these signs
for these five types of pause, example 1 may be represented as 3:
3.
cOn.dro : jO.khon | Os.te : na.mi.lo || tO.kho.no : ro.Ye.che | ra.ti,
pur.bo : di.ker | O.lOS : nO.Yo.ne || me.li.che : rOk.to | bha.ti
(cf. no. 1)
4.1.1.2.
Pause-based metrical units
Various
metrical units may be specified in terms of the above-mentioned 5 types of pause:
4.
i) The metrical unit specified by a pair of grade 5 pauses is called the verse
/ poNkti, e.g. cOndro jOkhon Oste namilo tOkhono roYeche rati of 1.
ii)
The metrical unit specified by a pair of grade 4 pauses is called the clause /
pOd, e.g. cOndro jOkhon Oste namilo of 1.
iii)
The metrical unit specified by a pair of grade 3 pauses is called the foot / pOrbo,
e.g. cOndro jOkhon of 1.
iv)
The metrical unit specified by a pair of grade 2 pauses is termed the sub-foot
/ upopOrbo, e.g. cOndro of 1.
v)
The metrical unit specified by a pair of grade 1 pauses is called the syllable
/ dOl e.g. cOn of 1.
For
terminological clarity, the pauses of grade 1, grade 2, grade 3, grade 4 and grade
5 may be called syllable pause, sub-foot pause, foot pause, caesural pause, and
verse pause respectively.
Three
notable points here are:
Firstly,
the pause immediately followed by the initial syllable of a verse is logically
equivalent to a verse pause, i.e. the strongest or grade 5 pause.
Secondly,
a metrical verse is not equivalent to a printed or a written line. In fact, depending
on its pause-divisions a single metrical verse may be arranged in more than one
line.
Thirdly, since
all the five grades of pause are hierarchically ordered, the overt existence of
any stronger grade pause logically ensures the covert existence of all its lower
grade pauses at that very location. Thus, although the statement clause is the
metrical division between a pair of grade 4 pauses, of 4ii, appears to be inexact,
actually it is not, because even a clause located at the beginning or end of a
verse is bound between a grade 4 pause and a pause stronger than grade 4, which
ensures the existence of another grade 4 pause.
This
logic holds good even in the cases of other metrical divisions, viz. foot, sub-foot,
and syllable.
4.1.1.3.
Location of pause
Among
the five grades of pause, only the location of grade 1 is categorically defined.
Grade 1 pause is located after each syllable, i.e. the minimum utterable unit
within a word.
Compared
to the location of grade 1 pause, those of grade 2 and grade 3 are not so well-defined.
Because unlike prose languages, where prose feet agree with sense pause, in verse
language metrical pause does not always coincide with sense pause.
Consequently, grade 2 and grade 3 pauses, generally having interword placement,
may even show intra-word placement, although within a word they are not placed
anywhere.
In fact,
the intra-word placement of grade 2 and grade 3 pauses obeys two principles, viz.
the Principle of Word Section / SObdopOrbo, and the Principle of Word Combination
/ joRa SObdo.
Principle
of Word Section: Small words in Bangla are composed of at most three syllables,
e.g. din 'day', jOn.mo 'birth', bO.Son.to 'spring' are monosyllabic, disyllabic,
and trisyllabic Bangla words respectively. But the words longer than these are
divided into groups containing 2 or 3 syllables each in actual speech. Such groups
may be termed as word sections, e.g. a.ra-dho.na 'worship', mo.ri-ci.ka 'a mirage',
O.po.ri-ci.to 'unknown', o.nu-kO.ro-ni.o 'exemplary' etc. Each of ara, dhona,
mori, Opori etc. is a word section.
Even the components of a compound word are considered as word sections, e.g. kali-daS
'servant of goddess Kali / a name', moha-bharot 'great India / name of an epic'
etc.
In pronunciation,
however, the weak pause between two word sections apparently is equivalent to
the grade 2 pause. Hence a grade 2 pause may logically be placed between two such
word sections.
Moreover,
often a grade 3 pause is placed between two word sections, i.e. often an intra-word
grade 2 pause is considered as a grade 3 pause.
Principle
of Word Combination: Unless there is some special semantic or metrical reason,
monosyllabic words, when spoken, are usually combined with the words preceding
or following them, e.g. ke-ke jabe? "who who go-will" 'Who all will
go?'; tumi-ki jabe? "you what go-will" 'Will you go?'; tumi na-gele
amio jabo-na "you not go-cond. I-too go-shall not" 'Unless you go I
too won't go' etc. Here ke-ke, tumi-ki, na-gele, and jabo-na represent word combinations.
In terms of metrical analysis, each of these combined words consisting of two
or three syllables may be considered as an undivided word, thus a sub-foot.
In brief, grade 2 and grade 3 pauses may be placed, firstly, at interword positions,
i.e. at word boundaries; and secondly, on the basis of the above two principles,
at intra-word positions, i.e. between word sections, e.g.
5.
jOl Se : cito | alo : bale || torun : SOho | -kare,
prio
: namTi | SikhiYe : dito || Sadher : Sari | -kare….
morbo : na-bhay | nipu : nika || cotu : rikar | Soke,
tara : SObay | onno : name || achen : mOrto | loke
(Sekal/ konika/ Tagore)2
"water
sprinkled around-root-loc. young mango-tree-loc.
dear name-cl. teach-conj.-gave
choice-gen. female-parrot-obj.
die-shall not-friend Nipunika Choturika-gen.
grief-loc.
they all different name-loc. are mortal-world-loc."
'They used to water the mango tree and teach their favourite female parrot the
beloved name.
Friends, I shall not die grieving for Nipunika and Chaturika,
(as I know that) they all exist in this mortal world under other names.'
In 5, in two words (viz. nipunika and coturika), grade 2 pause is placed between
word sections; in three words (viz. SOhokare, Sarikare, and mOrtoloke), grade
3 pause is placed between word sections. The first four words are undivided words,
whereas the fifth word is a compound word. na-bhay is a sub-foot of combined words.
The sub-foot divisions of the forms jOl Secito appear to be problematic. This
may be solved in two steps as follows:
In
the first step, in conformity to the Principle of Word Combination, monosyllabic
jOl is combined with the following trisyllabic word Secito, forming a tetrasyllabic
form jOlSecito. In the second step, by the Principle of Word Section, this tetrasyllabic
form is divided into two disyllabic word sections, (viz. jOl.Se and ci.to) and
the grade 2 pause is placed between these two word sections.
Now
I shall discuss the grade 4 and grade 5 pauses. For the sake of convenience I
shall discuss the grade 5 pause first.
The
placement of the grade 5 pause is more or less non-problematic - except for blank
verse, such pauses are placed verse-finally.
The
grade 4 pause is placed within a long verse, generally after two or three feet,
the last foot being left incomplete. In example 1, grade 4 pause is placed after
two complete feet and the second foot of the second half is left incomplete.
The
total number of the grade 4 pauses depends on the length of the metrical verse.
In a lengthy verse there may be more than one grade 4 pause, whereas in a considerably
small verse containing two or three feet there may be no grade 4 pause al all.
Between feet, there may be one or two grade 3 pauses, and after grade 3, instead
of grade 4, there comes the grade 5 pause in such small verses, e.g.
6.
kalo rati | gElo ghuce
alo tare | dilo muche
(2nd lesson/ SOhojpaTh 1/ Tagore)3
"black
night went finish-conj.
light it-obj. gave wipe-conj."
'The dark night vanished
(as) the light wiped it out'
In 6 each verse has two feet, divided by a grade 3 pause, and followed by a grade
5 pause.
The shortest
metrical verse in Bangla contains a single foot composed of two sub-feet divided
by a grade 2 pause and the entire foot is followed by a grade 5 pause, instead
of a grade 4 one. In other words, in this type the foot is synonymous with the
verse, e.g.
7. baYu :
bOY
bono : mOY
baMS : gach
kOre : nac
(1st lesson/ SOhojpaTh 1/ Tagore)4
"wind
blows
forest-wide
bamboo trees
do dance"
On the basis of the above discussion we may conclude that, for metrical analysis,
the obligatory units are the syllable, the foot, and the verse, whereas the sub-foot
and the clause may be regarded as optional, as they are often non-existent, as
in the case of short verses.
4.1.1.4. Elision of pause
In recitation sometimes it is found that the expected pause at some specific place
fails to appear. Such pause-loss is termed as elision of pause or pause skipping.
Elision of pause, however, comes under rule exceptions. In Bangla verse all the
five grades of pause may be elided as follows:
Elision
of grade 5 pause: Except for blank verse, grade 5 pause is never elided, because
the elision of such a pause would damage the basic metrical structure of a verse.
Elision of grade 4,
grade 3, and grade 2 pauses: Regarding these three types of elision of pause,
one general statement could be made: elision becomes imperative for these three
types of pause when there is either no word boundary or no suitable word section
at the places specified for these three types of pause. The reason behind this
are firstly, as we have already discussed, there can not be any intra-word grade
3 or grade 2 pause unless there is a word section; and secondly, in terms of the
hierarchical ordering of pauses, the elision of grade 2 and grade 3 pauses naturally
entails the elision of grade 4 pause too, e.g.
8.
kal bOle | ami sriSTi || kori ey | bhObo,
ghoRi
bOle | tahole a X mio sroSTa | tObo.
(upolokkho/ konika/ Tagore)5
"time
says I create do this material world
clock says then I-incl. creator your"
'Time says, "I am the creator of this material world",
the clock
says, "Then I am your creator".'
In 8, word section is non-existent at the place indicated by the sign [X], that
is why at this place the expected grade 3 pause (and thus also the grade 4 pause)
is elided.
9. sthOle
jO=le ar | gOgone : gOgone
baMSi ba=je jEno | modhur : lOgone
(SOrot/ kOlpona/ Tagore)6
"land-loc.
water-loc. and sky-loc. sky-loc.
flute plays as-if pleasant moment-loc."
'As
if the flute is being played everywhere - on land, in the water, in the air -
at pleasant moments'
In 9, no word section exists at the places indicated by the sign [=], as a result
of which the grade 2 pause is elided here.
Elision
of grade 1 pause: According to Sen, the elision of grade 1 pause gives rise to
a few metrical styles, for example,
9.
Su-r . jo cO ¦ le-n dhi . re || So-n . nE . Si | be . Se
po-S
. ci-m | no . di . ti . re || So-n . dha-r | de . Se
(tOpoSSa/ cittrobicittro/ Tagore)7
"sun
goes-hon. slowly ascetic dress-loc.
western river-bank-loc. dusk-gen. land-loc."
'In his ascetic dress the sun proceeds slowly
towards the land of dusk on
the western shore'
In 10, grade 1 pause is elided at the places indicated by the sign [-]. [¦]
indicates the elision of grade 3 pause.
One notable point here is that the elision of grade 1 pause, in my opinion, is
controversial because in this regard Sen's own statement seems to be contradictory.
On the one hand, Sen (1986: 10) decides to call the portion between a pair of
grade 1 pauses syllable. But on the other hand, the places where Sen (1986: 23)
indicates the elision of grade 1 pause by the sign [-], in terms of his own definition
of syllable (Sen, 1986: 14), are just devoid of any syllable boundary, i.e. at
such places the existence of any grade 1 pause is not even expected.
However, in the present chapter, I prefer to report the indigenous metrical study
as it is rather than to argue against it.
4.1.2.
Stress (proSSOr)
It
has already been mentioned that a foot starts with a stress accent and ends in
a pause. We have discussed pause. Now we shall deal with stress.
In Bangla verse metre, two grades of stress are available, which may be named
as the primary stress and the secondary stress. These two are hierarchically related.
In verse language,
the first sub-foot of a foot (and thus the foot itself) begins with a primary
stress, while the other sub-feet begin with secondary stress. In case the grade
3 pause (i.e. the foot pause) is located within a long word between two word sections,
then the initial syllable of the word section immediately following the grade
3 pause gets the primary stress; while the initial syllable of the initial word
section preceding the grade 3 pause gets the secondary stress. In other words,
under certain circumstances the primary stress may be placed in the middle of
a word, and consequently the secondary stress is placed at the beginning of that
word, e.g.
11. 1morbo
: 2na-bhay | 1nipu : 2nika || 1cotu : 2rikar | 1Soke,
1tara : 2SObay | 1onno : 2name || 1achen : 2mOrto | 1loke
(cf. no. 5)
In 11, each
initial sub-foot within a foot starts with a primary stress, indicated as 1; while
each second sub-foot starts with a secondary stress, indicated by 2. The verse
final feet are incomplete; they each consist of a single sub-foot. Both of these
incomplete sub-feet, viz. Soke and loke, begin with a primary stress. The initial
word section, viz. mOrto, of the word mOrto-loke provides the second sub-foot
for the preceding foot, and thus necessarily begins with a secondary stress. The
final word section, viz. loke, of mOrto-loke forms the initial sub-foot of the
next foot and consequently begins with a primary stress.
Like pause, stress too may be elided under certain circumstances. Usually the
stress indicating the commencement of a foot, immediately following a grade 4
or a grade 3 pause that is lost, is also lost, e.g.
12.
|dhire dhire | |SOrbori || |hOY Obo | |San,
|uThilo bi ¦ hONger || |prottuS | |gan.
(utSOb/ cittrobicittro/ Tagore)8
"slowly
slowly night becomes terminated
arose bird-gen. early-morning song"
'Gradually the dark night comes to an end and
there emerges the bird's chirping
at the day-break.'
In
12, the first grade 3 pause of the second verse is elided (indicated by the sign
[¦]), so is the stress (indicated by the absence of any stress mark) immediately
following it and indicating the starting point of the next foot. As a result of
the elision of the grade 3 pause and the stress here we get a clause without any
foot division.
Even
in 8 the foot-initial stress, viz. on mi of the word amio, immediately after the
lost grade 4 pause is elided - as a result of which this verse here is devoid
of any clause division.
On the basis of the above discussion of pause and stress we may come to the following
conclusions:
i) The syllable
is the minimal metrical unit.
ii)
The verse is the maximal metrical unit.
iii)
The existence of the sub-foot and the clause depends on the length of a verse.
iv)
Hence, between the syllable and the verse the foot is the most necessary and well-defined
metrical unit as it is clearly determined in terms of two fundamental factors,
viz. stress and pause.
4.2.
Syllable
A foot consists
of syllables, and a syllable is the minimum sound unit of a sentence utterable
by a single natural exertion of the organs of speech, e.g. dOl 'syllable', chOn.do
'metre', and chan.do.Sik 'metricist' - these three words, as they consist of one,
two, and three syllables respectively, are termed as monosyllabic, disyllabic,
and trisyllabic words respectively.
Before determining the relevance of syllable in the metrical analysis, it is necessary
to introduce the Bangla vowels as they are the basic components of the syllable.
4.2.1. Bangla vowels
Unlike Sanskrit, Bangla does not show any vowel length at its phonological level.
In pronunciation, all the Bangla vowels tend to be short, though under certain
circumstances they may even be long.
Bangla has six essential vowels, viz. O, a, i, u, e, and o, one of the basic components
of Bangla metre. Linguists prefer to include E in this list of Bangla vowels,
but in fact, E is nothing but an open variant of the vowel e. The difference between
these two vowels, viz. e and E, however, does not count in case of the unit-quantification
in metre.
These six
vowels are the independent complete vowels of Bangla, i.e. they occur in isolation.
Among these six vowels
i, u, e, and o have their dependent variants too. These dependent variants occur
only when they are sheltered by some preceding independent vowel, e.g. the sound
y, Y, w, and W (representing the dependent variants of the vowels i, e, u, and
o respectively) of the words koy 'where', khaY '(s)he eats', Dhew 'wave', jaW
'you go' etc. occur neither as independent vowels nor in isolation.
In the metrical terminology, such sheltered vowels are termed as dependent vowels
or vowel particles.
Bangla shows at least 16 combined vowels formed by combining these six independent
and four dependent vowels, e.g.
13.
oy : koy 'where', Soy 'friend (female)'
ow : bow 'wife', mow 'honey'
OY
: pOY.la '1st'
OW: hOW 'you be'
ay : nay 'not have'
aY : aY.na 'mirror'
aW : jaW 'you go' etc.
The
concepts of dependent vowel or vowel particle and combined vowels of metrical
field fully resemble those of the semivowels, viz. y, Y, w, and W, and diphthongs
of Ray et al. (1966), Dasgupta (1980), Sarkar (1985-86) of Bangla phonology respectively.
In metrical terminology
these six independent vowels, viz. O, a, i, u, e, and o, and the combined vowels,
i.e. oy, ow, OY, ay etc., are termed as open vowels and closed vowels respectively.
In other words, an independent vowel with an open provision of sheltering a vowel
particle immediately after it is called an open vowel; whereas a vowel for which
such provision has already been exercised as it has already given shelter to a
vowel particle after it is called a closed vowel. For example, in aY.na, in na,
the final syllable, a is an open vowel as it can shelter a vowel particle; whereas
aY, the initial syllable, is a closed vowel as it has already given shelter to
a vowel particle, viz. Y.
A
high degree of recurrence of these closed vowel sequences, on the one hand, constitutes
one of the chief phonological characteristics of Bangla; and on the other hand,
enriches the field of Bangla metrics.
4.2.2.
Dependent segments
It
has been shown that in Bangla i, u, e, and o generally occur as independent vowels,
though under specific circumstances their dependent occurrence is not infrequent.
On the contrary, the consonants are inherently dependent segments, i.e. the consonants
like t, n, p etc. cannot occur without either a preceding or a following vowel,
e.g. kO.tha 'speech', Ol.po 'little' etc.
Both the letters anusvara and visarga, the phonetic descriptions of which are
velar nasal and glottal fricative respectively, of Bangla script are inherently
dependent consonants which are always preceded, rather than followed, by vowels,
e.g. rON 'colour', bEN 'frog', uh/ ah 'expressions of annoyance/ pain' etc.
In Bangla script the symbol ( Ä ), called the has mark, nullifies the inherent
a of every consonant grapheme and turns it into a pure consonant, e.g. ßÄ,
àÄ, áÄ etc. denote /k, kh, g/ respectively. Sen (1986)
extends the use of this has symbol even to the vowel particles to indicate that
like pure consonants such particle too depend on independent vowels.
In brief, in Bangla the four dependent vowels and all the consonants may be termed
as dependent segments. In other words, dependent segments in Bangla are of two
types, viz. dependent vowels and dependent consonants. All of these dependent
segments may be described as sheltered segments.
4.2.3.
Syllable structure
Each
syllable consists of one and only one open or closed vowel, i.e. the obligatory
component of a syllable is one open or closed vowel, neither more nor less than
that. Both the types of vowel may be accompanied by one or more consonants, preceding
or following, e.g.
Syllable
consisting of vowel only
14.
o.Si 'sword'
a.ta 'custard apple'
oy.ra.bOt 'elephant' etc.
The initial syllables of the examples of 14 consist of vowels only. Such syllables
generally occur word initially and are not very frequent.
Vowel
+ consonant
15. og.ni
'fire'
al.ta 'a kind of cosmetic'
Ek.la 'alone' etc.
The initial syllables of the examples of 15 are to be considered. Such syllables
too, like the previous category, generally occur word initially and are not so
frequent.
Consonant +
open vowel
16. ke 'who'
tu.mi 'you'
Su.po.ri.ci.to 'well-known' etc.
Syllables of this type have the highest frequency in the language.
Consonant
+ vowel + consonant
17.
jOl 'water'
din 'day'
rat 'night' etc.
Like the previous type this type too is quite frequent.
Vowel
preceded by consonant cluster
18.
klan.ti 'fatigue'
stri 'wife' etc.
Vowel
followed by consonant cluster
19.
SONS.kri.to 'Sanskrit'
bhOrt.So.na 'scolding'
bENk 'bank' etc.
Syllables of this type are not available in Bangla proper; they are available
in either non-naturalized loan words from Sanskrit or English.
4.2.4.
Syllable type
In Bangla,
like vowels, syllables too are of chief two types, viz. open syllables and closed
syllables. A syllable not consisting of any dependent segment finally is an open
syllable; whereas a syllable which ends in one or more dependent segments is a
closed syllable, e.g. in a.kaS 'sky' the initial and final syllables are open
and closed respectively.
Though further sub-classification of open and closed syllable is not impossible,
for the purpose of metrical analysis these two types, viz. open and closed syllables,
are sufficient.
4.2.5.
Number of syllable
The
number of syllables in a word may be determined in two ways:
i)
By counting the total number of exertions by the organs of speech made during
the utterance of a word. The number of exertions is equal to the number of syllables.
But this method is hard to use as one may not have reliable intuitions about such
exertions.
ii) By counting
the total number of vowels, both open and closed, in a word. The number of vowels
is equal to the number of syllables because each syllable contains one and only
one vowel, e.g. kow.tuk 'fun', SaWM.tal 'Santhal' - these two words contain two
vowels and thus two syllables each; whereas the word paW.na.dar 'a creditor' contain
three vowels as well as three syllables. This method is quite easy to apply.
4.2.6.
Pronunciation variation of syllable
Though length is not a distinctive feature of Bangla vowels, length differentiation
is phonetically present in Bangla pronunciation. Hence, a discussion of Bangla
metrics must consider the length of the whole syllable, rather than that of the
vowel. In terms of the variations in pronunciation Bangla syllables are of chiefly
two types, viz. short and long.
Bangla open syllables are usually pronounced as short syllables. But the open
syllables that are connected neither to the preceding nor to the following syllable,
i.e. the open syllables in isolation, are normally pronounced as long syllables,
e.g.
20. coli coli |
pa- pa- | Toli Toli | jaY,
gOrobini
| heSe heSe | aRe aRe | caY.
(haSiraSi/ koRi o komol/ Tagore)9
"walking
walking step step tottering tottering goes,
proud-fem. smile-conj. smile-conj.
obliquely obliquely looks"
'The proud little girl smiles and glances sidelong as she totters step by step.'
In 20, the pronunciation of the two isolated monosyllables, viz. pa and pa, is
long.
Either for metrical
purposes or for the purpose of semantic clarity, Bangla closed syllables are often
separated or elongated in pronunciation. Thus they become long, i.e. each of them
becomes equal to two open syllables.
4.2.7.
Units for measuring syllable quantity
The
unit used to quantify a verse is a clause; a clause is measured in terms of feet.
Each foot, in turn, is a combination of syllables. Hence each foot is quantified
in terms of syllables.
In
Bangla metre, syllables are measured in terms of two types of unit, as follows:
i)
In a certain metrical variety, each syllable (irrespective of whether it is closed
or open) is considered as a single unit. Thus in this variety each foot is measured
in terms of what may be called the syllable unit.
ii)
In another variety the closed syllable is considered to have twice the metrical
quantity of an open syllable; and the quantity of sound equivalent to a short
open syllable is technically called a mora / kOla. Thus in this variety short
and long syllables are considered as monomoric and dimoric respectively. In this
variety, then, a foot is measured in terms of what may be called the moric unit.
In short, two types
of quantity-unit operate in the field of Bangla metre, viz. the syllabic unit
and the moric unit. Though both the units are based on the concept of syllable,
they belong to two different levels of structure.
4.3.
Three metrical patterns of Bangla
The auditory effect of a metrical style depends on the principles behind the distribution
or arrangements of its quantity-units. In Bangla such arrangements follow three
basic principles that result in three different metrical patterns as follows:
metrical patterns
______________|______________
syllabic
pattern / moric pattern
dOlobritto _____________|_____________
simple
moric / mixed moric /
SOrol kOlabritto composite /
missro kOlabritto
4.3.1.
Syllabic pattern
In
this variety, the metrical foot is formed in terms of the syllabic unit, i.e.
each syllable is conventionally considered as one unit of a foot in this pattern.
In other words, the
syllabic pattern is the metrical pattern in which the constituents of a foot are
the syllabic units, e.g.
21. | | | | | | | | | | | | | |
thakurdadar |
moto bone | achen riSi | muni,
| | | | | | | | | | | | | |
taMder pae
| pronam kore | gOlpo Onek | Suni.
(bonobaS/ SiSu/ Tagore)10
"grandfather-gen.
like forest-loc. are sage hermit,
their feet-loc. bend-and-touch do-conj.
stories many listen-I"
'In
the forests there are many sages and hermits as senior as grandfather. I pay homage
by bending and touching their feet and listen to a good number of stories from
them.'
In 21 each verse
consists of three complete and one incomplete feet. The complete feet consist
of four syllables; whereas the incomplete ones have two syllables each. The quantity
unit per syllable is indicated by the superscript [|]. In our pronunciation, in
this pattern, the closed syllables are shortened and equalled to the open syllables,
i.e. both the closed and the open syllables here are isochronous and thus considered
to be monomoric. As this pattern is based on the principle of uniformity of the
syllables, it is called the syllabic pattern.
Exceptions to this basic principle of uniformity of syllables are also evident
in this pattern. Such exceptions generally comprise the cases where in a verse
at some specific position some closed syllable is lengthened and made equal to
two open syllables, e.g. verse finally as follows:
22. | | | | | | | | | |
| | |
haY re kObe | keTe gEche | kalidaSer | kal,
| | | | | | | | | |
| | |
ponDitera | bibad kOre | loe tarikh | Sal,
| | | | | | | | | | |
| | |
apatoto | ey anonde | gOrbe bERay | nece
| | | | | | | | | | |
| | |
kalidaS to | namey achen | ami achi | beMce.
(Sekal/ konika/ Tagore)11
"alas
when cut-and-went Kalidas-gen. time,
learned-people dispute do regarding
date and year,
for-the-time-being this joy-loc. pride-loc. roam-around-I dancing
Kalidas-part. name-loc.-emp. is-hon. I am alive."
'Alas!
So long ago the time of Kalidas passed away that the learned people dispute regarding
the exact date and year of that. For the time being let me dance around and be
happy and proud being satisfied with the fact that Kalidas exists only in name
but I exist in life.'
The auditory judgement suggests that in 22, the two monosyllabic words, viz. kal
and Sal, consisting of one closed syllable each, are treated as equivalent to
the two disyllabic words, viz. nece and beMce, consisting of two open syllables
each, respectively.
4.3.2.
Simple moric pattern
In
this variety, the metrical foot is formed in terms of the moric unit.
In this simple moric style all the closed syllables are lengthened. Thus, in this
pattern the open and the closed syllables count as monomoric and dimoric respectively
in the process of foot quantification. Given the simplicity of such a process
of quantification, this pattern is named as the simple moric pattern, e.g.
23.
| | | | | | | | | | | | | |
kO-l . lo . le | ko . la . hO . le | ja . ge E-k
| dho . ni,
| | | | | | | | | | | | | |
O-n . dhe-r | kO-n . The-r | ga-n
a . go .| mo . ni.
(agomoni/ cittrobicittro/ Tagore)12
"a-roaring-wave-loc. a-loud-confused-noise-loc.
arises one melody
blind-gen. voice-gen. song Agomoni"13
'Amidst
the uproaring wave of the loud confused noises there emerges the tune of Agomoni
song in the blind man's voice.'
In 23, each verse contains four feet - three complete and one incomplete. Though
all the complete feet do not have four syllables, in our pronunciation each closed
syllable is lengthened and made equivalent to two open syllables. Lengthening
of closed syllables is indicated by the sign [-] in 23. The segment of speech
equivalent to a short open syllable is technically termed as mora. Hence in 23
if the open and the closed syllables count as monomoric and dimoric respectively
then we get four moras in each complete foot and two moras in each incomplete
one. In accordance with the quantity-unit, viz. moric unit, this pattern is named
as the simple moric pattern, in short the moric pattern.
Exceptions to this pattern usually appear in examples where closed syllables are
sometimes treated as monomoric, e.g.
24. | | | | | | | | | | | | | | | | |
jotsnaTuku | miSaY, baYu | dolaY keSo | paS,
| | | | | | | | | | | | | | |
| |
ekhoni tObe | probhat hObe | jagibe roSSiM | bhaS
(jotsnaloke/ benu o bina/ Satyendranath Dutta)14
"moonlight-little becomes-lost, wind waves locks-of
hair,
now-emp. then dawn be-will wake-will ray-glow"
'The
little moonlight becomes faint, the wind waves the locks of hair, shortly it will
be dawn and the glow of sun-beam will be there.'
In 24 the initial closed syllable of roSSiM, viz. roS, is shortened and treated
as monomoric.
4.3.3.
Mixed moric / composite pattern
In this variety too the metrical foot is formed in terms of the moric unit. This
branch of moric style, however, has its own characteristic pattern of distribution
and quantification of sounds. In this pattern each open syllable counts as monomoric;
whereas not all closed syllables count as dimoric. Word finally, closed syllables
are lengthened, but elsewhere they are not. As a consequence of this the closed
syllables, in this pattern, count as dimoric at that specific position, whereas
at other positions they count as monomoric. Because of this mixture of these two
types of closed syllable mora count, viz. the monomoric count, valid in the syllabic
pattern, and the dimoric count, valid in the simple moric pattern, this pattern
is termed as the mixed moric or the composite pattern, e.g.
25. | | | | |
| | | | | | | | |
bhalo mOndo | dukkho Su-kh | Ondhoka-r | alo,
| | | |
| | | | | | | | | |
mone hO-Y | SO-b nie | e dhOroni | bhalo.
(dhOratOl/
coytali/ Tagore)15
"good,
bad, sorrow, joy, darkness, light,
seem happens all including this world
splendid"
'The world
appears splendid with all its good and bad, sorrow and joy, darkness and light.'
In 25 each verse consists of four feet, three complete and one incomplete. As
is already mentioned, here the quantification of closed syllables varies according
to the pronunciation and position. The word final closed syllables are lengthened
and considered as dimoric. The closed syllable of a monosyllabic word too is considered
as a word final syllable. And the non-final closed syllables are usually shortened
and considered as monomoric.
According to the above calculation the three closed syllables in 25, viz. mOn,
duk, and On in the words mOndo, dukkho, and Ondhokar respectively, are treated
as monomoric, whereas the four closed syllables, viz. Sukh, kar, hOY, and SOb,
are treated as dimoric. The open syllables are naturally monomoric. In other words,
both the open syllables and the non-final closed syllables count as monomoric,
whereas the word final closed syllables count as dimoric. Thus the complete and
incomplete feet in 25 respectively contain four and two moras each. In this pattern
too the quantity-unit is the moric unit and hence it is treated as a special variety
under the moric pattern in which closed syllables show two different types of
count. Because of this characteristic feature it is called the mixed moric or
the composite pattern, in short, the mixed pattern.
Sen (1974) preferred to name this pattern as okkhorbritto / the letter-dependent
pattern. In his own words the justification behind this is that (Sen, 1974: 174)
'initially this pattern originated from the habit of counting okkhor / letters
while composing a foot…hence, calling this pattern as okkhorbritto is quite
justifiable.'
At the
initial stage of the development of this pattern, since the visual judgement was
considered as more important than the auditory judgement, there used to arise
confusions in two specific domains, viz. the domains of diphthongs and consonant
sequence, as follows:
Diphthongs:
Diphthongs are auditorily single units but visually they are sometimes single
but mostly double units - a fact that results in discrepancy between the auditory
and the visual judgements, e.g. in terms of okkhorbritto ¿úëÂ׿ù
/ Siwli 'a flower' consists of two moras but three letters; Ü / oy 'that'
consists of two moras but one letter.
Consonant
sequences: Likewise the consonant sequences of the non-tatsama words are not written
with the conventional conjunct consonantal letters, e.g. ý±ùÄß±
/halka 'light', ÂóúÄù± / pOSla 'a short
spell' etc. Here too the discrepancy is that both the examples are auditorily
dimoric, but visually they consist of three letters each.
Such discrepancies, however, bring about two types of effect in the field of Bangla
metrics, viz. metrical lapse and alternative spelling. The poetic works by Tagore
reveal such alternative spellings, e.g. Tagore used the letter Ü /oy 'that'
in both the syllabic and the moric styles (Sen, 1974: 166); but he preferred Ýý×
/oy 'that' for the composite style (Sen, 1974: 167). That is, he considered the
auditory judgement to be the primary one and attempted to make the visual judgement
agree with it.
However,
the exceptions to this composite pattern usually comprise four types of deviations
as follows:
i) In a
few instances even the word final closed syllables are treated as short, thus
monomoric.
ii) On the
contrary, in a few instances the non-final closed syllables are treated as long,
thus dimoric.
iii) The
final closed syllable of the first component of a tatsama compound word may oscillate
between two counts, viz. monomoric and dimoric, e.g. the closed syllable hrit
of the word hritpinDo 'heart' may be used either as a monomoric or as a dimoric
syllable.
iv) The closed
syllable between two such words that are combined with each other in terms of
both articulation and semantics may have an alternative monomoric treatment, e.g.
the word <ks?an?e ks?an?e> 'at short intervals' is pronounced as khOnekkhOne.
The closed syllable nek, which actually is the final syllable of the first component,
may count as monomoric in the composite pattern.
4.3.4.
Bangla poetry vs. the three metrical patterns of Bangla
Presently, Bangla poetry, or in other words, Bangla metrics is approximately a
thousand years old. For the purpose of metrical analysis it is convenient to divide
this thousand year time span into two chronological periods, viz. the pre-Tagore
period and the Tagore period.
The metrical literature of the pre-Tagore period includes cOrjapOd of the 10th/11th
cent.; srikriSnokirtOn; srikriSno bijOY; moNgol kabbo; dhamali; the literary works
by the poets Bharat Chandra and Ramprasad Sen of the 18th cent.; all the literary
works until the middle of the 19th cent., and some literary pieces belonging to
the 2nd half of the 19th cent.
Analyses reveal that the metrical patterns of this period belong to two chief
categories, viz. those evolved from some Old Indo-Aryan or Middle Indo-Aryan metrical
style, and those representing the preliminary stage of the three metrical patterns
of Bangla. Between these two categories the first one, apparently, is not so relevant
in the discussion of Bangla metre; whereas the analyses of second type provides
a few salient pieces of information, as follows:
i)
The composite style was first initiated in the verses of srikriSnokirtOn of the
15th cent.
ii) In the
written literature the syllabic style was first used consciously in the verses
of dhamali by Lochandas of the 16th cent. The syllabic style in the pre-dhamali
literature was hardly available as this particular style was considered as the
folk metre and was restricted to the oral folk literature.
iii)
The moric pattern was first employed in the songs of goddess Kali by poet Ramprasad
Sen of the 18th cent.
The
three metrical patterns of Bangla flourished most during the Tagore period. These
three patterns, viz. the syllabic, the moric and the composite, found their deep-rooted
foundation in the works of Tagore, (and thus obviously in the Bangla literature)
through successful and creative experiments. The total number of Tagore's books
of verse is 60, and the number of poem is 2949. Among these 2949, 1216 represent
the composite pattern, 849 represent the moric pattern, 760 represent the syllabic
pattern, and 124 represent prose verse, a style in which the poet very skilfully
and successfully superseded the conventional metrical patterns.
Observation
shows that until 1890 the composite pattern was predominant in Tagore's poems.
As a poet-metricist Tagore correctly felt that if the long vowels of Sanskrit
are treated as long, thus dimoric, then they sound clearly artificial in Bangla.
But if instead the closed syllables are treated as dimoric they do agree with
the nature of Bangla pronunciation. As the consequence of this feeling the poet
in his 1890 book of verse called manoSi treated the closed syllables as dimoric.
Hence in the history of Bangla metrics manoSi is considered as a prominent landmark.
In the 1887 poem bhulbhaNa, included in manoSi, the moric style, which treats
the closed syllables as dimoric, was first employed very successfully and efficiently.
Tagore employed the
syllabic pattern, the so called folk pattern, in serious poetic works in his book
of verse called khonika of 1900.
Not only in metrical patterns but in foot-formation etc. also Tagore carried out
extensive experiments.
The immediately pre and post Tagore period shows a good number of excellent poetic
works which very successfully restructured the conventional metrical patterns
of Bangla. Among these non-conventional trends mention could be made of the works
of Michael Madhusudan Dutta, of the pre-Tagore period and those of Jibanananda,
Sudhindranath Dutta, Buddhadeb Basu, Bishnu De, Subhash Mukhopadhyay, Sankha Ghose,
Nirendranath Chakraborty, Shakti Chattopadhyay, Sunil Gangopadhyay, Kabita Sinha,
Joy Goswami, Mahbub Ul Alam Choudhury, Shamsur Rahman and so on belonging to the
post-Tagore period.
4.4.
Conclusion
The phonological
characterization of Bangla metrics has been accomplished in the above discussion.
Apart from this phonological level the complete metrical analysis of Bangla by
Sen (1986) includes a few other levels too, viz. the level of rhyme, and the structural
level of sub-foot, foot, clause, verse, and stanza. A cover term for all these
structural levels by Sen is the morphology of Bengali versification.
However
interesting the discussion of the morphology of Bengali versification may be,
it is apparently not quite relevant for the purpose of the present book, oriented
as it is to metrical phonology as a tool for analysing language. Hence at this
point I leave these levels untouched.
The
present chapter has served to illustrate the methodology of metrical analysis
in terms of indigenous, Bangla-based research traditions. In the next chapter
we shall attempt to look at some phonological aspects of Bangla in terms of the
current western theories of metrical phonology.
NOTES
1.
Cited from Sen, 1986, pp. 9.
2.
ibid, pp. 14.
3. ibid,
pp. 17.
4. ibid, pp.
18.
5. ibid, pp. 20.
6.
ibid, pp. 22.
7. ibid,
pp. 24.
8. ibid, pp.
21.
9. ibid, pp. 29.
10.
ibid, pp. 32.
11. ibid,
pp. 36.
12. ibid, pp.
32.
13. agomoni: the
songs about the coming of Uma, the wife of Shiva, to her father's house - as is
told in the Hindu legends.
14.
Cited from Sen, 1986, pp. 39.
15.
ibid, pp. 32.