Please look at the wonderful page by Richard Ishida for more information. Note also that the current page is not about the bengali language, it is about the script. A little information about the language is also available elsewhere. There is also a page describing how the script arose.
In this page, I do not use unicode, rather, I use embedded images so that it can be viewed everywhere. The symbol appears whenever the script I was using did not have the letter to demonstrate its use. (The current unicode scripts are complete, but I haven't updated this page yet.)
Apart from punctuations, digits, and other symbols used to write numbers, the Bengali script is an alphabetic script, but like all the Indian scripts, is better understood as a syllabic script where the symbols for the syllables are often, though not always, derived from the constituent sounds of the syllable. For the isolated consonant-vowel and consonant-vowel-consonant syllables, this makes it into an ‘abiguda’, but this is supplemented by a lot of rules for consonant-consonant ligatures. The abiguda nature of the script derives its origins from the brAhmi script of north India dating back to a few centuries before the common era. Though it is considered certain that the concept of alphabetic writing came to India from the aramaic script, the script seems to be an invention in India. With a study of phonetics being part of saMskRta grammatical studies, a field in which Indians developed logic just as Euclid did with geometry in Greece, it is not surprising that the basic alphabet is organized in roughly logical principles. Though, here I describe Bengali writing, I borrow heavily from saMskRta terminology as needed.
The primary division in the alphabet is between vowels and consonants. An open syllable consists of a series of consonants followed by a vowel, whereas a closed syllable ends in a series of consonants following an open syllable. The script is designed to depict open syllables: closed syllables are treated as an open syllable followed by an incomplete one.
Thus the letters depicting the vowels are used when they form the sole component of a syllable. There are the following vowels a, A, i, I, u, U, R, e, ai, o and au. (an i or an u is written as _i or _u if the previous syllable ends in an a). a, i, u and R are short vowels, the rest are long, though in modern Bengali all take on a short value. a and A are guttural vowels, i, I, e and ai are palatal, u, U, o and au are labial and R is retroflex. The first part of the alphabet till R is closed and frontal, the rest are back and open. ai and au are dipthongs, and in modern bengali are pronounced like o+i (IPA [oj]) and o+u (IPA [ow]). The vowel e often takes on the value of the English a in cat (IPA [æ] if you can see the unicode character), though sometimes it keeps its original value (IPA [a]). The value of a is extremely variable: in initial syllable it often sounds like a short version of the the vowel in English caught (i.e. the IPA [ɔ] if you can see unicode), and in the final syllable of a word-unit it is often omitted unless following multiple consonants or standing alone. In other positions, it is usually replaceable by o (IPA [o]) in pronounciation. In Bengali, R is pronounced as rhi (IPA [ri] is common), and the saMskRta vowels RR (long version of R), L (short dental closed) and LL are not used, though L used to be counted in the alphabet till recently. They do have symbols ( respectively), but they are not used in writing Bengali. The A is an open back unrounded vowel (IPA [α]). The rest of the vowels (i, I, u, U, e and o) are illustrated in the English words hit, heat, put, root, late and cold, though since length is suppressed in bengali, the IPA symbols would be [i] for the first two, [u] for the next two, and [e] and [o] for the last. The superlong form of the long vowels that exist in saMskRta are missing in Bengali.
In addition, there is a vowel symbol a^yA () which does not appear in the alphabet. It is technically the single syllable ayA (as opposed to the syllable yA following the syllable a), and is pronounced like the a in English cat (IPA [æ]). In ancient bengali, a sometimes took the place of Y, and even had vowel modifiers (see below) attached to it. These obolete forms are represented as a^i, a^I etc. (Note that in the script, a^a and a^A cannot exist, they would be the same as a or A.)
There is also a symbol which indicates the letter a which is assimilated into a previous e or o by the rules of liasion. It is used only when writing saMskRta in the bengali script.
There are two special characters, M and H, which counts as being vowels and consonants. These are pronounced like ng (IPA [ŋ]) and a hard breathing like h (IPA [h] but voiceless) at the end of a syllable, and hence these always stand alone, never combined with the succeeding vowels or consonants. If M is followed by a closed consonant, its pronounciation changes to be the same as the fifth of the group to which the following consonant belongs (see description of closed consonants below). In counting the alphabet, these are conventionally counted at the end of the vowels in the form aM and aH; of course they do occur with any vowel whatsoever preceding them whatsoever. Including these, the vowels are then:
a
A
i
I
u
U
R
e
ai
o
au
aM
aH
The symbol for M is called anusvar, and the one for H is caled visarga. saMskRta also had two breathing symbols (H.) which were considered half an H, one (which occurred before the labial consonants) was softer than the other (which typicaly appeared before the guttural consonants). These have symbols (thus aH. is ), but they are not used when writing bengali. In addition, there is a symbol called candravindu that nasalizes the vowel in any syllable, thus one can modify a to ã (IPA [ɔ] to [ɔ̃]) (). Note that õ standing alone as a word is, completely illogicaly, pronounced as oM (IPA [oŋ]) (or, rarely, om, IPA [om]). In addition, the modifier can stand by itself (which I represent by ~) in front of a name indicating that the person is no longer alive: in this context, it is called Ishvara (IPA [iʃʃɔr]).
The consonants are divided into the closed consonants (called vargya) and open consonants (called antastha). The former come in five groups of five and in each group represent an unvoiced (aghoSa) unaspirated (alpaprANa), unvoiced aspirated (mahAprANa), voiced (ghoSa) unaspirated, voiced aspirated and voiced unaspirated nasal (nAsikya) respectively. The basic symbol for the consonant is illustrated using the symbol for a syllable consisting of that consonant and a. The five groups are the guttural (called kaNThya or ka vargIYa: k, kh, g, gh, G ) (IPA [k], [kʰ], [g], [gʱ], [ŋ]), palatal (called tAlavya or ta vargIYa: c, ch, j, jh, and J ) (IPA [tʃ], [tʃʰ], [dʒ], [dʒʱ], [ɲ]), retroflex (called murdhanya or Ta vargIYa: T, Th, D, Dh, N ) (IPA [ʈ], [ʈʰ], [ɖ], [ɖʱ], [n]), dental (called dantya or ta vargIYa: t, th, d, dh, n ) (IPA [t̪], [t̪ʰ], [d̪], [d̪ʱ], n]), and labial (called oSThya or pa vargIYa: p, ph, b, bh, m ) (IPA [p], [pʰ], [b], [bʱ], [m]) consonants; their pronounciations are like the English counterparts except the following: G is pronounced like ng, c like ch in English church, J like ñ of spanish, T, Th, D, Dh and N are slightly harder than in English whereas t, th, d, and dh are slightly softer as in French. In modern Bengali, the difference between N (murdhanya Na) and n (dantya na) is hardly noticeable.
The open consonant consist of the semivowels (y/Y, r, l, v ) and the sibilants (sh, S, s, h ). The palatal semivowel (kaNThatAlavya) y is now pronounced like j (IPA [dʒ]) (and called antastha ja) except in combinations with a variant Y pronounced like the english y (in IPA it becomes either [e] or disappears) (and called antastha a, in old bengali texts, a plain vowel is often used instead of Y), and the labial semivowel (dantyoSThya) v (called antastha va) indistinguishably from b (IPA [b]) (called vargya ba). The retroflex (murdhanya) r (IPA [r]) is more trilled than in English and the dental l is like in English (IPA [l]). The retroflex (murdhanya) S, dental (dantya) s and palatal (tAlavya) sh sibilants are all pronounced almost equivalently in bengali (IPA [ʃ] allophonic with [s]) like the initial consonant in English sugar; and h (written as _h in rare cases like in the syllable containing b + h; as needed in old bengali usage) is pronounced like the non-silent English h (IPA [h]). There are also the rolled forms D. ( called Dae shunyo D.a) and Dh. ( called Dhae shunyo Dh.a) which are much harder than r (IPA [ɽ] and [ɽʱ].
These follow the general pattern illustrated below for the consonant k:
k
ka
kA
ki
kI
ku
kU kR kRR
kL
kLL
ke
kai
ko
kau
(If needed, I represent the the isolated vowel modifier as ^^i etc. The vowel modifier for a^yA is usually just written and treated as yA, but in the rare case when confusion is possible, it is written as ^yA.) However, the vowels u and U combines with the consonant r and the consonant clusters tr, str, bhr, and shr to the side as illustrated by ru and rU (). Furthermore, gu, shu, hu and hR are special ().
Consonant clusters are the hardest part of the Bengali script to learn. A simple rule is that the conjoined consonants are to be read from top to bottom and from left to right. However, a number of consonants change when combined. Thus, the combined forms of m-, r-, l- S- and s- are exemplified by the following:
mba
rka
lka
STa ska
Similarly, the combined forms of -t, -th -dh, -n, -T, -m, -y, -r, -l, -v/-b and -s are illustrated below:
pta
stha
ddha
nna
STa dma
kya
dra
kla
dva
ksa
(Note that -yA is pronounced like the vowel in English cat, otherwise addition of -y or -v to any consonant other than r duplicates the sound of that consonant.)
There are, in addition, a lot of special forms which are not recognizably related to the constituents. The following provides a list of the difficult ones (Jcha should be added to the list):
kka
kTa
kta
kra
kSa
Gka
Gkha
Gga
jja
jJa
Jca
Jja
TTa
tta
ttha
tra
dda
nta
ntra
nya
ppa
bhra
STha
SNa
sTa
sya
hma
hNa
Note that kSa is pronounced kkha and jJ as gy.
In Bengali whole numbers are written using the decimal place value system. The symbols for the digits are:
0
1
2
3
4
5
6
7
8
9
In addition, there are special symbols to indicate simple fractions: the following were supposed to be the representation of the fractions 1/16 thru 15/16: , except that the font I am using for this has really messed it up! That 3/4 business seems upside down, and the 2/16 and 3/16 are so badly written that they seem tilted versions of the same. In addition whole amounts are postfixed by the symbol , and the currency symbol (a form of the first letter of the word TAkA) is .
Their are two native bengali punctuation symbols (transliteration: | and ||, Bengali: ) which indicate respectively the end of a sentence or a line of poetry, and a heavier break (like the end of a couplet or stanza in poetry). In practice, the punctuation marks of the roman script, such as comma, question mark, and quotation marks and others are also used.