Bengali script: Wikis


Note: Many of our articles have direct quotes from sources you can cite, within the Wikipedia article! This article doesn't yet, but we're working on it! See more info or our list of citable articles.

Did you know ...

More interesting facts on Bengali script

Include this on your site/blog:


From Wikipedia, the free encyclopedia

This article contains Indic text. Without proper rendering support, you may see question marks or boxes, misplaced vowels or missing conjuncts instead of Indic text.
Bengali abugida
Type Abugida
Spoken languages Bengali
Time period 11th Century to the present [1]
Parent systems
ISO 15924 Beng
Note: This page may contain IPA phonetic symbols in Unicode.

The Bengali script (Bengali: বাংলা লিপি Bangla lipi) is the writing system for the Bengali language. It is also used for Assamese, Bishnupriya Manipuri, Kokborok, Garo and Mundari languages. All these languages are spoken in the eastern region of South Asia. Historically, the script has also been used to write the Sanskrit language in the same region. From a classificatory point of view, the Bengali script is an abugida, i.e. its vowel graphemes are mainly realized not as independent letters like in a true alphabet, but as diacritics attached to its consonant graphemes. It is written from left to right, lacks distinct letter cases, and is recognizable by a distinctive horizontal line running along the tops of the letters that links them together. It is less blocky and presents a more sinuous shaping than Devanagari, another popular abugida used in India and Nepal.



The Bengali script evolved from the Eastern Nagari script, which belongs to the Brahmic family of scripts, along with the Devanagari script and other written systems of the Indian subcontinent. Both Eastern Nagari and Devanagari were derived from the ancient Nagari script. In addition to differences in how the letters are pronounced in the different languages, there are some minor typographical differences between the version of the script used for Assamese and Bishnupriya Manipuri, and that used for Bengali and other languages. For example, the letter (Bengali র; Assamese ৰ; Bishnupriya Manipuri র/ৰ) and (Bengali second ব; Assamese/Bishnupriya Manipuri ৱ) have distinct variations depending on the language being written.

The Bengali script was originally not associated with any particular language, but was often used in the eastern regions of Medieval India. It was standardized and modernized by Ishwar Chandra under the auspices of the British East India Company. The script was originally used to write Sanskrit, which for centuries was the only written language of the Indian subcontinent in addition to Tamil. Epics of Hindu scripture, including the Mahabharata or Ramayana, were written in older versions of the Bengali script in this region. After the medieval period, the use of Sanskrit as the sole written language gave way to Pali, and eventually to the vernacular languages we know now as Bengali and Assamese. Srimanta Sankardeva used it in the 15th and 16th centuries to compose his oeuvre in Assamese and Brajavali the language of the Bhakti poets. It was also used by the later Ahom kings to write the Buranjis, the Ahom chronicles, in the Assamese language. There is a rich legacy of Indian literature written in this script, which is still occasionally used to write Sanskrit today.


In the Bengali script, clusters of consonants are represented by different and sometimes quite irregular characters; thus, learning to read the script is complicated by the sheer size of the full set of characters and character combinations, numbering about 350. While efforts at standardizing the script for the Bengali language continue in such notable centers as the Bangla Academies (unaffiliated) at Dhaka (Bangladesh) and Kolkata (West Bengal, India), it is still not quite uniform as yet, as many people continue to use various archaic forms of letters, resulting in concurrent forms for the same sounds. Among the various regional variations within this script, only the Assamese and Bengali variations exist today in the formalized system.

It seems likely that the standardization of the script will be greatly influenced by the need to typeset it on computers. The large alphabet can be represented, with a great deal of ingenuity, within the ASCII character set, omitting certain irregular conjuncts. Work has been underway since around 2001 to develop Unicode fonts, and it seems likely that it will split into two variants, traditional and modern.

In this and other articles on Wikipedia dealing with the Bengali language, a Romanization scheme used by linguists specializing in Bengali phonology is included along with IPA transcription.

A recent effort by the government of West Bengal focused on simplifying bengali spellings in primary school texts.

Description of Bengali graphemes

The inventory of the Bengali script is made up of a finite number of graphemes, i.e. fundamental units of a written language. These can be divided into vowel graphemes, consonant graphemes (including consonant conjuncts), modifier graphemes, digits and punctuation marks.

Vowel graphemes

The Bengali script has a total of 11 vowel graphemes, each of which is called a স্বরবর্ণ shôrobôrno "vowel letter". These shôrobôrnos represent six of the seven main vowel sounds of Bengali, along with two vowel diphthongs. All of these are used in both Bengali and Assamese, the two main languages using the script. There is no standard character in the script for the Bengali main vowel sound /æ/, and vowel length differences thought to be represented by different vowel graphemes (e.g., hrôshsho i vs. dirgho i) do not hold true for the spoken language. Also, the grapheme called ri does not really represent a vowel phoneme, rather the sound /ri/.

All Bengali vowel graphemes (except one) have two allographic forms. When a vowel sound is at the beginning of a syllable or when it follows another vowel, it is written using the "full form" (or independent form) of the vowel grapheme. But when a vowel follows a consonant (or a consonant cluster, which behaves as a single unit for the purpose of vowel sign placement), the vowel grapheme is written in a shortened "diacritic form" (or dependent form) which, depending on the vowel, can appear above, below, before or after the consonant sign. The diacritic form of a vowel grapheme cannot appear without a consonant. A diacritic form is named by adding a "-kar" to the end of the name of the corresponding vowel grapheme (see table below).

An exception to the above system is the grapheme for the vowel /ɔ/. This grapheme has no diacritic form, but is considered inherent in every consonant grapheme. To specifically denote the absence of this inherent vowel [ɔ] attached to a consonant, a diacritic called the hôshonto (্) may be written underneath the consonant.

Although there are only two diphthong graphemes in the inventory of the script, the Bengali sound system has in fact many diphthongs.[2] Most of these diphthongs are represented by juxtaposing the graphemes of their forming vowels, as in কেউ keu /keu/.

The table below shows the vowel graphemes present in the modern (i.e., since late nineteenth century) inventory of the Bengali script, which has abandoned the graphemes for three historical vowels, rri, li, and lli, traditionally placed between ri and e.

Bengali vowel graphemes (স্বরবর্ণ shôrobôrno)
Full form Name of
full form
Diacritic form,
used with
the consonant [kɔ] (ক)
Name of
diacritic form
Transliteration IPA
shôro ô
(shôre ô)
"vowel ô"
/ɔ/ and /o/ [3] (none) (none) kô and ko /kɔ/ and /ko/
shôro a
(shôre a)
"vowel a"
/a/ কা akar ka /ka/
hrôshsho i
(hrôshsho i)
"short i"
/i/ কি hrôshsho ikar
ki /ki/
dirgho i
"long i"
/i/ কী dirgho ikar
ki /ki/
hrôshsho u
(rôshsho u)
"short u"
/u/ কু hrôshsho ukar
ku /ku/
dirgho u
"long u"
/u/ কূ dirgho ukar
ku /ku/
ri /ri/ কৃ rikar/rifôla kri /kri/
e /e/ and /æ/ কে ekar kê and ke /ke/ and /kæ/
oi /oj/ কৈ oikar koi /koj/
o /o/ কো okar ko /ko/
ou /ow/ কৌ oukar kou /kow/

Consonant graphemes

Consonant graphemes are called ব্যঞ্জনবর্ণ bênjonbôrno "consonant letter" in Bengali. The names of these consonant graphemes are typically just the consonant's main pronunciation plus the inherent vowel ô. Since the inherent vowel is assumed and not written, most graphemes' names look identical to the grapheme itself (e.g. the name of the grapheme ঘ is itself ঘ ghô). Some graphemes that have lost their distinctive pronunciation in Modern Bengali are called by a more elaborate name. For example, since the consonant phoneme /n/ can be written ন, ণ, or ঞ (depending on the spelling of the particular word), these graphemes are not simply called ; instead, they are called দন্ত্য ন donto nô ("dental n"), মূর্ধন্য ণ murdhonno nô ("cerebral n"), and ঞীয়/ইঙ niô/ingô. Similarly, the phoneme /ʃ/ can be written as শ talobbo shô ("palatal s"), ষ murdhonno shô ("cerebral s"), or স donto shô ("dental s"), depending on the word. Since the consonant ঙ /ŋ/ cannot occur at the beginning of a word in Bengali, its name is not ঙ ngô but উঙ ungô (pronounced by some as উম umô or উঁঅ ũô). Similarly, since semivowels ([j], [w], [e̯], [o̯]) cannot occur at the beginning of a Bengali word, the name for "semi-vowel " য় is not অন্তঃস্থ য় ôntostho e̯ô but অন্তঃস্থ অ ôntostho ô.

In the earlier grapheme inventories of Bengali script, one can find a second (called ôntostho bô) following . This ôntostho bô originally represented a /v/ or /w/ sound, but later merged with the borgio bô in Bengali. The two bô's were represented with identical symbols occurring in two different places in the inventory. In the modern inventories of the Bengali script, only borgio bô is retained.

Bengali consonant graphemes (ব্যঞ্জনবর্ণ bênjonbôrno)
Letter Name of
consonant grapheme
Transliteration IPA
k /k/
khô kh /kʰ/
g /ɡ/
ghô gh /ɡʱ/
ungô, umô ņ /ŋ/
chô ch /tʃ/
chhô chh /tʃʰ/
borgio jô
j /dʒ/
jhô jh /dʒʱ/
ingô, niô n /n/
ţô ţ /ʈ/
ţhô ţh /ʈʰ/
đô đ /ɖ/
đhô đh /ɖʱ/
murdhonno nô
n /n/
t /t̪/
thô th /t̪ʰ/
d /d̪/
dhô dh /d̪ʱ/
donto nô
n /n/
p /p/
phô ph /pʰ/
b /b/
bhô bh /bʱ/
m /m/
ôntostho jô
j /dʒ/
(bôe shunno) rô r /ɾ/
l /l/
talobbo shô
sh and s /ʃ/ / /s/
murdhonno shô
(peţ kaţa shô)
sh /ʃ/
donto shô
sh and s /ʃ/ / /s/
h /h/
য় ôntostho ô
e and - /e̯/ /-
ড় đôe shunno/bindu ŗô ŗ /ɽ/
ঢ় đhôe shunno/bindu ŗô ŗh /ɽ/

Consonant conjuncts

Up to four consecutive consonants not separated by vowels can be orthographically represented as a ligature called a "consonant conjunct" (Bengali: যুক্তাক্ষর juktakkhor or যুক্তবর্ণ juktobôrno). Typically, the first consonant in the conjunct is shown above and/or to the left of the following consonants. Many consonants appear in an abbreviated or compressed form when serving as part of a conjunct. Others simply take exceptional forms in conjuncts, bearing little or no resemblance to the base character.

Often, consonant conjuncts are not actually pronounced as would be implied by the pronunciation of the individual components. For example, adding ল underneath শ shô in Bengali creates the conjunct শ্ল, which is not pronounced shlô but slô in Bengali. Many conjuncts represent Sanskrit sounds that were lost thousands of years before modern Bengali was ever spoken, as in জ্ঞ, which is a combination of জ and ঞ niô, but is not pronounced jnô. Instead, it is pronounced ggõ in Bengali. Thus, as conjuncts often represent (combinations of) sounds that cannot be easily understood from the components, the following descriptions are concerned only with the construction of the conjunct, and not the resulting pronunciation. Thus, a variant of the IAST romanization scheme is used instead of the phonemic romanization:

Fused forms

Some consonants fuse in such a way that one stroke of the first consonant also serves as a stroke of the next.

  • The consonants can be placed on top of one another, sharing their vertical line: ক্ক kkô গ্ন gnô গ্ল glô ন্ন nnô প্ন pnô প্প ppô ল্ল llô etc.
  • As the last member of a conjunct, ৱ wô and ব bô can hang on the vertical line under the preceding consonants, taking the shape of ব bô (here referred to as বফলা bôfôla): গ্ব gwô ণ্ব ṇwô দ্ব dwô/dbô ল্ব lwô শ্ব śwô.
  • The consonants can also be placed side-by-side, sharing their vertical line: দ্দ ddô ন্দ ndô ব্দ bdô ব্জ bjô প্ট pṭô শ্চ ścô শ্ছ śchô etc.

Approximated forms

Some consonants are simply written closer to one another to indicate that they are in a conjunct together.

  • As the last member of a conjunct, গ gô can appear unaltered, with the preceding consonant simply written closer to it: দ্গ dgô.
  • As the last member of a conjunct, ৱ wô and ব bô can appear immediately to the right of the preceding consonant, taking the shape of ব bô (here referred to as বফলা bôfôla): ধ্ব dhwô ব্ব bbô হ্ব hwô.

Compressed forms

Some consonants are compressed (and often simplified) when appearing as the first member of a conjunct.

  • As the first member of a conjunct, the consonants ঙ ŋô চ cô ড ḍô ব bô are often compressed and placed at the top-left of the following consonant, with little or no change to the basic shape: ঙ্ক্ষ ŋkṣô ঙ্খ ŋkhô ঙ্ঘ ŋghô ঙ্ম ŋmô চ্চ ccô চ্ছ cchô চ্ঞ cñô ড্ড ḍḍô ব্ব bbô.
  • As the first member of a conjunct, ত tô is compressed and placed above the following consonant, with little or no change to the basic shape: ত্ন tnô ত্ম tmô ত্ব twô.
  • As the first member of a conjunct, ম mô is compressed and simplified to a curved shape. It is placed above or to the top-left of the following consonant: ম্ন mnô ম্প mpô ম্ফ mfô ম্ব mbô ম্ভ mbhô ম্ম mmô ম্ল mlô.
  • As the first member of a conjunct, ষ ṣô is compressed and simplified to an oval shape with a diagonal stroke through it. It is placed to the top-left of the following consonants: ষ্ক ṣkô ষ্ট ṣṭô ষ্ঠ ṣṭhô ষ্প ṣpô ষ্ফ ṣfô ষ্ম ṣmô.
  • As the first member of a conjunct, স sô is compressed and simplified to a ribbon shape. It is placed above or to the top-left of the following consonant: স্ক skô স্খ skhô স্ট sṭô স্ত stô স্থ sthô স্ন snô স্প spô স্ফ sfô স্ব swô স্ম smô স্ল slô.

Abbreviated forms

Some consonants are abbreviated when appearing in conjuncts, losing part of their basic shape.

  • As the first member of a conjunct, জ jô can lose its final downstroke: জ্জ jjô জ্ঞ jñô জ্ব jwô.
  • As the first member of a conjunct, ঞ ñô can lose its bottom half: ঞ্চ ñcô ঞ্ছ ñchô ঞ্জ ñjô ঞ্ঝ ñjhô.
  • As the last member of a conjunct, ঞ ñô can lose its left half (the এ part): জ্ঞ jñô.
  • As the first member of a conjunct, ণ ṇô and প pô can lose their downstroke: ণ্ঠ ṇṭhô ণ্ড ṇḍô প্ত ptô প্স psô.
  • As the first member of a conjunct, ত tô and ভ bhô can lose their final upward tail: ত্ত ttô ত্থ tthô ত্র trô ভ্র bhrô.
  • As the last member of a conjunct, থ thô can lose its final upstroke, taking the form of হ hô instead: ন্থ nthô ম্থ mthô স্থ sthô.
  • As the last member of a conjunct, ম mô can lose its initial downstroke: ক্ম kmô গ্ম gmô ঙ্ম ŋmô ট্ম ṭmô ণ্ম ṇmô ত্ম tmô দ্ম dmô ন্ম nmô ম্ম mmô শ্ম śmô ষ্ম ṣmô স্ম smô.
  • As the last member of a conjunct, স sô can lose its top half: ক্স ksô ন্স nsô.

Variant forms

Some consonants have forms that are used regularly, but only within conjuncts.

  • As the first member of a conjunct, ঙ ŋô can appear as a loop and curl: ঙ্ক ŋkô ঙ্গ ŋgô.
  • As the last member of a conjunct, the curled top of ধ dhô is replaced by a straight downstroke to the right: গ্ধ gdhô দ্ধ ddhô ন্ধ ndhô ব্ধ bdhô.
  • As the first member of a conjunct, র rô appears as a diagonal stroke (called রেফ ref) above the following member: র্ক rkô র্খ rkhô র্গ rgô র্ঘ rghô etc.
  • As the last member of a conjunct, র rô appears as a wavy horizontal line (called রফলা rôfôla) under the previous member: খ্র khrô গ্র grô ঘ্র ghrô ব্র brô etc.
    • In some fonts, certain conjuncts with রফলা rôfôla appear using the compressed (and often simplified) form of the previous consonant: জ্র jrô ট্র ṭrô ঠ্র ṭhrô ড্র ḍrô ম্র mrô স্র srô.
    • In some fonts, certain conjuncts with রফলা rôfôla appear using the abbreviated form of the previous consonant: ক্র krô ত্র trô ভ্র bhrô
  • As the last member of a conjunct, য yô appears as a wavy vertical line (called যফলা jôfôla) to the right of the previous member: ক্য kyô খ্য khyô গ্য gyô ঘ্য ghyô etc.
    • In some fonts, certain conjuncts with যফলা jôfôla appear using special fused forms: দ্য dyô ন্য nyô শ্য śyô ষ্য ṣyô স্য syô হ্য hyô.


  • When followed by র rô, ক kô takes on the abbreviated form of ত tô with the addition of a curl to the right: ক্র krô.
  • When preceded by the abbreviated form of ঞ ñô, চ cô takes the shape of ব bô: ঞ্চ ñcô
  • When preceded by another ট ṭô, ট ṭô is reduced to a leftward curl: ট্ট ṭṭô.
  • When preceded by ষ ṣô, ণ ṇô appears as two loops to the right: ষ্ণ ṣṇô.
  • As the first member of a conjunct, or when word-final and followed by no vowel, ত tô can appear as ৎ (called খণ্ড-ত khônḍo tô or "broken "): ৎস tsô ৎপ tpô ৎক tkô etc.
  • When preceded by হ hô, ন nô appears as a curl to the right: হ্ন hnô.
  • Certain combinations simply must be memorized: ক্ষ kṣô হ্ম hmô.

Exceptional consonant-vowel combinations

When serving as a vowel sign, উ u, ঊ ū, and ঋ take on many exceptional forms.

  • উ u
    • When following গ gô or শ śô, it takes on a variant form resembling the final tail of ও: গু gu শু śu.
    • When following a ত tô that is already part of a conjunct with ন nô or স sô, it is fused with the ত tô to resemble ও o: ন্তু ntu স্তু stu.
    • When following র rô, and in many fonts also following the variant রফলা rôfôla, it appears as an upward curl to the right of the preceding consonant as opposed to a downward loop below: রু ru গ্রু gru ত্রু tru থ্রু thru দ্রু dru ধ্রু dhru ব্রু bru ভ্রু bhru শ্রু śru.
    • When following হ hô, it appears as an extra curl: হু hu.
  • ঊ ū
    • When following র rô, and in many fonts also following the variant রফলা rôfôla, it appears as a downstroke to the right of the preceding consonant as opposed to a downward hook below: রূ rū গ্রূ grū থ্রূ thrū দ্রূ drū ধ্রূ dhrū ভ্রূ bhrū শ্রূ śrū.
    • When following হ hô, it takes the variant shape of ঊ ū: হৃ hṛ.

Conjuncts of three consonants also exist, and follow the same rules as above. Examples include স sô + ত tô +র rô = স্ত্র strô, ম mô + প pô + র rô = ম্প্র mprô, ঙ ŋô + ক kô + ষ ṣô = ঙ্ক্ষ ŋkṣô, জ jô + জ jô + ৱ wô = জ্জ্ব jjwô, ক kô + ষ ṣô + ম mô = ক্ষ্ম kṣmô. Theoretically, four-consonant conjuncts can also be created, as in র rô + স sô + ট ṭô + র rô = র্স্ট্র rsṭrô, but these are not found in real words.

Modifiers and others

Modifier and other graphemes in Bengali
Symbol with [kɔ] (ক) Name Function Transliteration IPA
ক্ hôshonto
"final hôsh"
Suppresses the inherent vowel [ɔ] k /k/
কৎ khônđo tô
"broken tô"
Final unaspirated dental [t̪] (ত) kôt /kɔt̪/
কং ônushshôr Final velar nasal kôņ /kɔŋ/
কঃ bishôrgo
  1. pronounced as syllable-final voiceless breath, as in উঃ
  2. isn't pronounced, but geminates the following consonant, as in দুঃসময়
  3. isn't pronounced at all, as in দুঃস্থ


  1. [uh]
  2. [d̪uʃːomɔj]
  3. [d̪ustʰo]
কঁ chôndrobindu
Vowel nasalization kôñ /kɔ̃/

ঃ and ং are also often used as abbreviation marks in Bengali, like the full stop in English. For example ডঃ "Dr." stands for ডক্টর "Doctor" and নং "no." stands for নম্বর "number". In modern Bengali usage, however, the full stop is gaining in currency as the abbreviation mark.

The jôphôla is sometimes used as a diacritic to indicate non-Bengali vowels of various kinds in transliterated foreign words. For example, the schwa is indicated by a jôphôla, the French u and the German umlaut ü as উ্য, the German umlaut ö as ও্য or এ্য, etc.

The apostrophe, known in Bengali as ঊর্ধ্বকমা urdhokôma "upper comma", is sometimes used to distinguish between homographs, as in পাটা paţa "plank" and পা'টা paţa "the leg". Sometimes a hyphen is used for the same purpose (as in পা-টা, an alternative of পা'টা).

Digits and numerals

The Bengali script has ten digits (graphemes or symbols indicating the numbers from 0 to 9), which are variants of Indian numerals (known as Arabic numerals in the West). Bengali digits have no horizontal headstroke or "matra".

Bengali digits
English digits 0 1 2 3 4 5 6 7 8 9
Bengali digits
Bengali names
of digits
shunno êk dui tin char pãch chhôe shat nôe
শূন্য এক দুই তিন চার পাঁচ ছয় সাত আট নয়

Numbers larger than 9 are written in Bengali using a positional base 10 numeral system (the decimal system), just as in English. A period or dot is used to denote the decimal separator, which separates the integral and the fractional parts of a decimal number. When writing large numbers with many digits, commas are used as delimiters to group digits, indicating the thousand (হাজার hajar), the hundred thousand or lakh (লাখ lakh or লক্ষ lokkho), and the ten million or hundred lakh or crore (কোটি koṭi or ক্রোড় kroṛ) units. In other words, going leftwards from the decimal separator, the first grouping consists of three digits, and the subsequent groupings always consist of 2 digits.

For example, the English number 17,557,345 will be written in traditional Bengali as ১,৭৫,৫৭,৩৪৫ (এক কোটি, পঁচাত্তর লাখ, সাতান্ন হাজার, তিন শ পঁয়তাল্লিশ êk koṭi pãchattor lakh, shatanno hajar, tin sho pôetallish, "one crore, seventy-five lakhs, fifty-seven thousand, three hundred forty-five").

The matra

Whereas in western scripts (Latin, Cyrillic, etc.) the letter-forms stand on an invisible baseline, the Bengali letter-forms hang from a visible horizontal left-to-right headstroke called মাত্রা matra (not to be confused with its Hindi cognate matra, which denotes the dependent forms of Hindi vowels). The presence and absence of this matra can be important. For example, the letter ত [tɔ] and the numeral ৩ "3" are distinguishable only by the presence or absence of the matra, as is the case between the consonant cluster ত্র [trɔ] and the independent vowel এ [e]. The letter-forms also employ the concepts of letter-width and letter-height (the vertical space between the visible matra and an invisible baseline).

Punctuation marks

Bengali punctuation marks, apart from the downstroke daŗi (|), the Bengali equivalent of a full stop, have been adopted from western scripts and their usage is similar. Commas, semicolons, colons, quotation marks, etc. are the same as in English. The concept of using capital letters is absent in the Bengali script, hence proper names are unmarked.

Characteristics of the Bengali text


Bengali text is written and read horizontally, from left to right.


The consonant graphemes and the full form of vowel graphemes fit into an imaginary rectangle of uniform size (i.e. uniform width and height). The size of a consonant conjunct, regardless of its complexity, is deliberately maintained the same as that of a single consonant grapheme, so that diacritic vowel forms can be attached to it without any distortion.


In a typical Bengali text, orthographic words, i.e., words as they are written, can be seen as being separated from each other by an even spacing. Graphemes within a word are also evenly spaced, but this spacing is much narrower than the spacing between words.

Punctuation marks

As discussed earlier, the Bengali punctuation marks are often the same as their English counterparts, both in form and function.

Characteristics of the orthographic word

In every Bengali orthographic word, one can find different kinds of graphemes and their combinations, and they are as follows:

  • Independent form of vowel graphemes, which can be found at the beginning of a word or after another vowel sound.
  • Consonant graphemes (or consonant conjuncts) with no diacritic vowel form attached to them. An inherent vowel (either /ɔ/ or /o/, depending on context[3]) is nevertheless assumed to be attached to them.
  • Consonant graphemes (or consonant conjuncts) with a diacritic vowel form attached to them.
  • Other modifier symbols indicating nasalization of vowels, suppression of the inherent vowel, etc.

The "matra"

The matra or the horizontal headstroke on each grapheme usually add up and often form a continuous single headstroke over the entire orthographic word, with different graphemes hanging down from it. This gives Bengali text a distinct look. Among other modern Indic scripts, Devanagari also has this characteristic.


The following inconsistencies are inherent in the Bengali script and orthography. They often put additional burden on the person learning the script. The inconsistencies manifest themselves in various ways. Sometimes there are multiple different letters or symbols for the same sound (over-production). Sometimes a letter loses its original sound value. In other instances, the coverage of phonological information by the script is incomplete, inconsistent and/or ambiguous. Most of these inconsistencies can be attributed to the fact that the script was originally conceived to represent Sanskrit sounds.

Redundant graphemes for the vowel sounds [i] and [u]

The Bengali script has two symbols for the vowel sound [i] and two symbols for the vowel sound [u]. This redundancy stems from the time when this script was used to write Sanskrit, a language that had a short [i] and a long [iː], and a short [u] and a long [uː]. These letters are preserved in the Bengali script with their traditional names of hrôshsho i/u (lit. "short i/u") and dirgho i/u (lit. "long i/u") despite the fact that they are no longer pronounced differently in ordinary speech. These graphemes do serve an etymological function, however, in preserving the original Sanskrit spelling in tôtshomo Bengali words (i.e., words that were borrowed from Sanskrit).

The vowel grapheme ri

The grapheme called ri does not really represent a vowel phoneme in Bengali, rather the consonant-vowel combination /ri/. Nevertheless, it is included in the vowel section of the inventory of the Bengali script. This inconsistency is also a remnant from Sanskrit, where the grapheme represents a retroflex approximant and regarded as a vowel in Sanskrit.

The vowel sound [æ]

Even though the near-open front unrounded vowel [æ] is one of the seven main vowel sounds in the standard Bengali language, no distinct vowel symbol has been allotted for it in the script, since there is no [æ] sound in Sanskrit, the primary written language when the script was conceived. As a result, this sound is orthographically realized by multiple means in modern Bengali orthography, usually using some combination of এ, অ and the jôfôla (diacritic form of the consonant grapheme য ôntostho jô) as seen in the following examples:

  • word-initially: এত [æt̪o] "so much", এ্যাকাডেমী [ækademi] "academy", অ্যামিবা [æmiba] "amoeba"
  • word-medially and following a consonant: দেখা [d̪ækha] "to see", ব্যস্ত [bæst̪o] "busy", ব্যাকরণ [bækɔron] "grammar".

ত and ৎ

In native or tôdbhôbo Bengali words, syllable-final ত is pronounced /t̪/, as in নাতনি /nat̪ni/ "grand daughter", করাত /kɔrat̪/ "saw", etc.

ৎ (called খণ্ড-ত khônḍo tô "broken ") is always used syllable-finally and always pronounced as /t̪/. It is predominantly found in loan words from Sanskrit such as ভবিষ্যৎ "future", সত্যজিৎ "Satyajit (a proper name)", etc. It is also found in some onomatopoeic words (such as থপাৎ "sound of something heavy that fell", মড়াৎ "sound of something breaking", etc.), as the first member of some consonant conjuncts (such as ৎস tsô, ৎপ tpô, ৎক tkô, etc.), and in some foreign loanwords (e.g. নাৎসি /nat̪si/ "Nazi", জুজুৎসু /dʒudʒut̪su/ "Jujutsu", etc.) which contain the same conjuncts.

This is an over-production inconsistency, where the sound /t̪/ is realized by both ত and ৎ. This creates confusion among inexperienced writers of Bengali. There is no simple way of telling which symbol should be used. Usually, the contexts where ৎ is used need to be memorized, as these are less frequent.

শ, ষ and স

Three graphemes—শ talobbo shô "palatal s", ষ murdhonno shô "cerebral s", and স donto shô "dental s"—are used to represent the voiceless palato-alveolar fricative [ʃ], as seen in their word-final pronunciations in ফিসফিস [pʰiʃpʰiʃ] "whisper", বিশ [biʃ] "twenty" and বিষ [biʃ] "poison". The grapheme স donto shô "dental s", however, does retain the voiceless alveolar fricative [s] sound when used as the first component in certain consonant conjuncts as in স্খলন [skʰɔlon] "fall", স্পন্দন [spɔndon] "beat", etc.

জ and য

There are two letters (জ and য) for the voiced postalveolar affricate [dʒ]. Compare জাল [dʒal] and যাও [dʒao].

What was once pronounced and written as a retroflex nasal ণ [ɳ] is now pronounced as an alveolar [n] (unless conjoined with another retroflex consonant such as ট, ঠ, ড and ঢ), although the spelling does not reflect this change.


The romanization of Bengali is the representation of the Bengali language in the Latin script. While different standards for romanization have been proposed for Bengali, these have not been adopted with the degree of uniformity seen in languages such as Japanese or Sanskrit[4]. Most standardized Bengali romanizations are adapted from standards proposed for Indic languages, and these models are compared below.

Transliteration vs Transcription

The Romanization of a language written in a non-Roman script can be based on transliteration (orthographically accurate, i.e. the original spelling can be recovered) or transcription (phonetically accurate, i.e. the pronunciation can be reproduced). This distinction is important in Bengali as its orthography was adopted from Sanskrit, and ignores sound change processes of several millennia. To some degree, all writing systems differ from the way the language is pronounced, but this may be more extreme for languages like Bengali. For example, the three letters শ, ষ, and স had distinct pronunciations in Sanskrit, but over several centuries, the standard pronunciation of Bengali (usually modeled on the Nadia dialect), has lost these phonetic distinctions (all three are usually pronounced as IPA [ʃ]) while the spelling distinction nevertheless persists in orthography.

In written texts, it is easy to distinguish between homophones such as শাপ shap "curse" and সাপ shap "snake". Such a distinction could be particularly relevant in searching for the term in an encyclopedia, for example. However, the fact that the words sound identical means that they would be transcribed identically; thus, some important meaning distinctions cannot be rendered in a transcription model. Another issue with transcription systems is that cross-dialectal and cross-register differences are widespread, and thus the same word or lexeme may have many different transcriptions. Even simple words like মন "mind" may be pronounced "mon", "môn", or (in poetry) "mônô" (e.g. the Indian national anthem, Jana Gana Mana).

Often, different phonemes (meaningfully different sounds) are represented by the same symbol or grapheme. Thus, the vowel এ can represent both [e] (এল elo [elo] "came"), or [ɛ] (এক êk [ɛk] "one"). Occasionally, words written in the same way (homographs) may have different pronunciations for differing meanings: মত can mean "opinion" (pronounced môt), or "similar to" (môto). Thus, some important phonemic distinctions cannot be rendered in a transliteration model. In addition, when representing a Bengali word to allow speakers of other languages to pronounce it easily, it may be better to use a transcription, which does not include the silent letters and other idiosyncrasies (e.g. স্বাস্থ্য shastho, spelled <swāsthya>, or অজ্ঞান ôggên, spelled <ajñāna>) that make Bengali orthography so complicated.

Comparison of Romanizations

Comparisons of standard romanization schemes for Bengali are given in the table below. Two standards are commonly used for transliteration of Indic languages including Bengali. Many standards (e.g. NLK / ISO), use diacritic marks and permit case markings for proper nouns. Newer forms (e.g. Harvard-Kyoto) are more suited for ASCII-derivative keyboards, and use upper- and lower-case letters contrastively and forgo normal standards for English capitalization.

  • "NLK" stands for the diacritic-based letter-to-letter transliteration schemes, best represented by the National Library at Kolkata romanization or the ISO 15919, or IAST. This is the ISO standard, and it uses diacritic marks (e.g. ā, ) to reflect the additional characters and sounds of Bengali letters.
  • ITRANS is an ASCII representation for Sanskrit; it is one-to-many, i.e. there may be more than one way of transliterating characters, which can make internet searching more complicated. ITRANS representations forgo capitalization norms of English so as to be able to represent the characters using a normal ASCII keyboard.
  • "HK" stands for two other case-sensitive letter-to-letter transliteration schemes: Harvard-Kyoto and XIAST scheme. These are similar to the ITRANS scheme, and use only one form for each character.
  • XHK or Extended Harvard-Kyoto (XHK) stands for the case-sensitive letter-to-letter Extended Harvard-Kyoto transliteration. This adds some specific characters for handling Bengali text to IAST.
  • "Wiki" stands for a phonemic transcription-based romanization. It is a sound-preserving transcription based on what is perceived to be the standard pronunciation of the Bengali words, with no reference to how it is written in Bengali script. It uses diacritics often used by linguists specializing in Bengali (other than IPA), and is the transcription system used to represent Bengali sounds in Wikipedia articles.


The following table includes examples of Bengali words Romanized using the various systems mentioned above.

Example words
In orthography Meaning NLK XHK ITRANS HK Wiki IPA
মন mind mana mana mana mana mon [mon]
সাপ snake sāpa sApa saapa sApa shap [ʃap]
শাপ curse śāpa zApa shaapa zApa shap [ʃap]
মত opinion mata mata mata mata môt [mɔt̪]
মত like mata mata mata mata môto [mɔt̪o]
তেল oil tēla tela tela tela tel [t̪el]
গেল went gēla gela gela gela gêlo [ɡɛlo]
জ্বর fever jvara jvara jvara jvara jôr [dʒɔr]
স্বাস্থ্য health svāsthya svAsthya svaasthya svAsthya shastho [ʃast̪ʰo]
বাংলাদেশ Bangladesh bāṃlādēśa bAMlAdeza baa.mlaadesha bAMlAdeza bangladesh [baŋlad̪eʃ]
ব্যঞ্জনধ্বনি consonant byañjanadhvani byaJjanadhvani bya~njanadhvani byaJjanadhvani bênjondhoni [bɛndʒond̪ʱoni]
আত্মহত্যা suicide ātmahatyā AtmahatyA aatmahatyaa AtmahatyA attõhotta [at̪ːõhot̪ːa]

Romanization Reference

The IPA (International Phonetic Alphabet) transcription is provided in the rightmost column, representing the most common pronunciation of the glyph in Standard Colloquial Bengali, alongside the various romanizations described above.

a a a a ô/o [ɔ]/[o]
ā ā A~aa A a [a]
i i i i i [i]
ī ī I~ii I i [i]
u u u u u [u]
ū ū U~uu U u [u]
RRi~R^i R ri [ri]
ē e e e ê/e [ɛ]/[e]
ai ai ai ai oi [oj]
ō o o o o [o]
au au au au ou [ow]
k k k k k [k]
kh kh kh kh kh [kʰ]
g g g g g [ɡ]
gh gh gh gh gh [ɡʱ]
~N G ng [ŋ]
c c ch c ch [tʃ]
ch ch Ch ch chh [tʃʰ]
j j j j j [dʒ]
jh jh jh jh jh [dʒʱ]
ñ ñ ~n J n [n]
T T ţ [ʈ]
ṭh ṭh Th Th ţh [ʈʰ]
D D đ [ɖ]
ড় .D P ŗ [ɽ]
ḍh ḍh Dh Dh đh [ɖʱ]
ঢ় ḍh ḏh .Dh Ph ŗ [ɽ]
N N n [n]
t t t t t [t̪]
th th th th th [t̪ʰ]
d d d d d [d̪]
dh dh dh dh dh [d̪ʱ]
n n n n n [n]
p p p p p [p]
ph ph ph ph ph [ph]
b b b b b [b]
bh bh bh bh bh [bʱ]
m m m m m [m]
y y y j [dʒ]
য় y Y Y e/- [e]/-
r r r r r [r]
l l l l l [l]
ś ś sh z sh/s [ʃ]/[s]
Sh S sh [ʃ]
s s s s sh/s [ʃ]/[s]
h h h h h [ɦ]
H H varies varies
.m M ng [ŋ]
ɱ .N ~ ~ [~]
্য y y y y varies varies
্ব v v v v varies varies
ক্ষ kṣ kṣ x kS kkh/kh [kʰː]/[kʰ]
জ্ঞ GY Jj gg/g [ɡː]/[ɡ]
শ্র śr śr shr zr sr [sr]

Bengali in Unicode

The Unicode range for Bengali is U+0980 ... U+09FF.

Bengali chart (PDF)
  0 1 2 3 4 5 6 7 8 9 A B C D E F
U+09Bx             ি

Grapheme frequency

According to Bengali linguist Munier Chowdhury, the following 9 graphemes are the most frequent in Bengali texts [5]:

Grapheme Percentage

Collating sequence

There is yet to be a uniform standard collating sequence (sorting order of graphemes to be used in dictionaries, indices, computer sorting programs, etc.) of Bengali graphemes. Experts in both India and Bangladesh are currently working towards a common solution for this problem.

Sample Texts

Article 1 of the Universal Declaration of Human Rights

Bengali in Eastern Nagari script

ধারা ১: সমস্ত মানুষ স্বাধীনভাবে সমান মর্যাদা এবং অধিকার নিয়ে জন্মগ্রহণ করে। তাঁদের বিবেক এবং বুদ্ধি আছে; সুতরাং সকলেরই একে অপরের প্রতি ভ্রাতৃত্বসুলভ মনোভাব নিয়ে আচরণ করা উচিৎ।

Bengali in Romanization

Dhara êk: Shômosto manush shadhinbhabe shôman môrjada ebong odhikar nie jônmogrohon kôre. Tãder bibek ebong buddhi achhe; shutorang shôkoleri êke ôporer proti bhrattrittoshulôbh monobhab nie achoron kôra uchit.

Bengali in IPA

d̪ʱara æk ʃɔmost̪o manuʃ ʃad̪ʱinbʱabe ʃɔman mɔrdʒad̪a eboŋ od̪ʱikar nie dʒɔnmoɡrohon kɔre. t̪ãd̪er bibek eboŋ bud̪ʱːi atʃʰe; ʃut̪oraŋ ʃɔkoleri æke ɔporer prot̪i bʱrat̪ːrit̪ːoʃulɔbʱ monobʱab nie atʃoron kɔra utʃit̪.


Clause 1: All human free-manner-in equal dignity and right taken birth-take do. Their reason and intelligence exist; therefore everyone-indeed one another's towards brotherhood-ly attitude taken conduct do should.


Article 1: All human beings are born free and equal in dignity and rights. They are endowed with reason and conscience. Therefore, they should act towards one another in a spirit of brotherhood.

Jana Gana Mana

The following is a sample text of script, from the song Jana Gana Mana (জন গণ মন Jôno Gôno Mono). The selection is a Bengali song, written in Shadhubhasha (সাধুভাষা) style. The song was later adopted as the national anthem of India. It was written by Rabindranath Tagore (রবীন্দ্রনাথ ঠাকুর Robindronath Ṭhakur) who is acknowledged as the single most important and defining figure of Bengali literature.

জনগণমন-অধিনায়ক জয় হে ভারতভাগ্যবিধাতা!
পঞ্জাব সিন্ধু গুজরাট মরাঠা দ্রাবিড় উত্কল বঙ্গ
বিন্ধ্য হিমাচল যমুনা গঙ্গা উচ্ছলজলধিতরঙ্গ
তব শুভ নামে জাগে, তব শুভ আশিস মাগে,
গাহে তব জয়গাথা।
জনগণমঙ্গলদায়ক জয় হে ভারতভাগ্যবিধাতা!
জয় হে, জয় হে, জয় হে, জয় জয় জয়, জয় হে ॥

In Romanization:

Jônogônomono-odhinaeoko jôeô he Bharotobhaggobidhata!
Pônjabo Shindhu Gujoraţo Môraţha Drabiŗo Utkôlo Bônggo,
Bindho Himachôlo Jomuna Gôngga Uchchhôlojôlodhitoronggo,
Tôbo shubho name jage, tôbo shubho ashish mage,
Gahe tôbo jôeogatha.
Jônogônomonggolodaeoko jôeô he Bharotobhaggobidhata!
Jôeo he, jôeo he, jôeo he, jôeo jôeo jôeo, jôeo he!

Footnotes and references

  1. ^ Ancient Scripts
  2. ^ Different Bengali linguists give different numbers of Bengali diphthongs in their works depending on methodology, e.g. 25 (Chatterji 1939: 40), 31 (Hai 1964), 45 (Ashraf and Ashraf 1966: 49), 28 (Kostic and Das 1972:6-7) and 17 (Sarkar 1987).
  3. ^ a b The natural pronunciation of the grapheme অ, whether in its independent (visible) form or in its "inherent" (invisible) form in a consonant grapheme, is /ɔ/. But its pronunciation changes to /o/ in the following contexts:
    • অ is in the first syllable and there is a ই /i/ or উ /u/ in the next syllable, as in অতি oti "much" /ot̪i/, বলছি bolcchi "(I am) speaking" /ˡboltʃʰi/
    • if the অ is the inherent vowel in a word-initial consonant cluster ending in rôfôla "rô ending" /r/, as in প্রথম prothom "first" /prot̪ʰom/
    • if the next consonant cluster contains a jôfôla "jô ending", as in অন্য onno "other" /onːo/, জন্য jonno "for" /dʒonːo/
  4. ^ In Japanese there exists some debate as to whether to accent certain distinctions, such as Tōhoku vs Tohoku. Sanskrit is well standardized, because the speaking community is relatively small, and sound change is not a large concern
  5. ^ See Chowdhury 1963


  • Ashraf, Syed Ali; Ashraf, Asia (1966), "Bengali Diphthongs", in Dil A. S., Shahidullah Presentation Volume, Lahore: Linguistic Research Group of Pakistan, pp. 47-52  
  • Chatterji, Suniti Kumar (1939), Vasha-prakash Bangala Vyakaran (A Grammar of the Bengali Language), Calcutta: University of Calcutta  
  • Chowdhury, Munier (1963), "Shahitto, shônkhatôtto o bhashatôtto (Literature, statistics and linguistics)", Bangla Academy Potrika (Dhaka) 6 (4): 65-76  
  • Kostic, Djordje; Das, Rhea S. (1972), A Short Outline of Bengali Phonetics, Calcutta: Statistical Publishing Company  
  • Hai, Muhammad Abdul (1964), Dhvani Vijnan O Bangla Dhvani-tattwa (Phonetics and Bengali Phonology), Dhaka: Bangla Academy  
  • Sarkar, Pabitra (1987), "Bangla Dishôrodhoni (Bengali Diphthongs)", Bhasha (Calcutta) 4-5: 10-12  

External links

digital encoding and rendering

Got something to say? Make a comment.
Your name
Your email address