Bengali Script

Bengali script (also called Bangla script) is one of the most widely used writing systems in the world, serving over 300 million people as the primary script for Bengali (the official language of Bangladesh and the Indian state of West Bengal) and Assamese. Descended from the ancient Brahmi script through Siddham and proto-Bengali forms, the modern Bengali script took shape around the 11th century CE. In Unicode, Bengali presents significant rendering challenges due to its complex conjunct consonants, vowel sign placement, and contextual shaping requirements. This guide explores the script's structure, its Unicode encoding, and the technical considerations for working with Bengali text.

History and Background

Bengali script belongs to the Eastern Nagari family of scripts, sharing ancestry with Assamese, Maithili, and Tirhuta scripts. Its evolution can be traced through several stages:

The script achieved its modern printed form through the work of Ishwar Chandra Vidyasagar and the Serampore Mission Press in the 19th century. The characteristic matra (the horizontal headline connecting letters, similar to Devanagari's shirorekha) became a standard typographic feature in printed Bengali.

Script Structure

Bengali is an abugida (alphasyllabary) where each consonant letter carries an inherent vowel /a/ (or /o/ in Bengali pronunciation). Other vowels are indicated by diacritical marks (vowel signs) added to the consonant.

Vowels

Bengali has 11 vowel letters (independent forms used at the start of words or syllables) and corresponding vowel signs (dependent forms attached to consonants):

Note the vowel signs for ো (o) and ৌ (au) — these are composite, appearing both to the left and right of the consonant simultaneously. In Unicode, these are encoded as two-part vowel signs: ে (U+09CB left part) + া (right part) for ো, forming a single visual unit around the consonant.

Consonants

Bengali has 35 consonant letters in the basic set:

Each consonant carries the inherent vowel /a/ (pronounced /o/ in standard Bengali). To suppress the inherent vowel (creating a "dead" consonant), the hasanta (virama, U+09CD) is used.

Conjunct Consonants (যুক্তাক্ষর)

One of Bengali script's most complex features is its system of conjunct consonants (juktakkhor) — ligatures formed when two or more consonants occur together without an intervening vowel. Rather than writing each consonant with a hasanta between them, Bengali typically merges them into a combined form:

ক + ্ + ষ → ক্ষ  (ksha — a single conjunct glyph)
স + ্ + ত → স্ত  (sta)
ন + ্ + ত → ন্ত  (nta)

Bengali has hundreds of conjunct forms, many of which look nothing like their component letters. This makes Bengali one of the most demanding scripts for font design — a comprehensive Bengali font must include glyphs for all common conjuncts, mapped through OpenType GSUB (Glyph Substitution) tables.

Some notable conjunct examples:

The Reph and Ya-phala

Two consonants have special combining behavior that appears throughout Bengali text:

• Reph: When র (ra) appears before another consonant with a hasanta, it takes a special form called reph — a small hook above the following consonant cluster: র + ্ + ক → র্ক (rka, with reph above ক)

• Ya-phala: When য (ya) appears after a consonant with a hasanta, it takes a subscript form called ya-phala (a curved stroke below the consonant): ক + ্ + য → ক্য (kya, with ya-phala below ক)

• Ra-phala: Similarly, র after a hasanta becomes a subscript diagonal stroke: ক + ্ + র → ক্র (kra, with ra-phala below ক)

The Unicode Bengali Block

The block is organized as follows:

Bengali Digits

Bengali has its own numeral system, though Arabic (Western) numerals are increasingly common:

Special Characters

• Chandrabindu (U+0981): Nasalization mark (ँ)
• Anusvara (U+0982): Nasal sound marker (ং)
• Visarga (U+0983): Aspiration marker (ঃ)
• Hasanta/Virama (U+09CD): Suppresses inherent vowel, triggers conjunct formation
• Khanda Ta (U+09CE): A special form of ত without inherent vowel, used word-finally
• Bengali Rupee Sign (U+09F3): ৳

Text Rendering Pipeline

Rendering Bengali text correctly requires a sophisticated shaping engine. The process involves multiple steps:

1. Character Reordering

Left-position vowel signs (ি, ে, ৈ) are stored after their consonant in Unicode (logical order) but rendered before it (visual order). The rendering engine must reorder these:

Stored:   ক (U+0995) + ি (U+09BF)
Rendered: কি  (the ি appears to the left of ক)

2. Conjunct Formation

When the engine encounters a consonant + hasanta + consonant sequence, it checks the font's GSUB table for a matching conjunct glyph:

Input:  ক (U+0995) + ্ (U+09CD) + ষ (U+09B7)
Lookup: GSUB table → conjunct glyph for ক্ষ
Output: Single conjunct glyph ক্ষ

If no conjunct glyph exists in the font, the hasanta is displayed explicitly.

3. Mark Positioning

Above-marks (chandrabindu, reph) and below-marks (vowel signs ু, ূ, ra-phala) are positioned using the font's GPOS (Glyph Positioning) table.

Common Rendering Issues

Working with Bengali in Code

Python

import unicodedata

# Bengali character properties
char = "\u0995"  # ক (ka)
print(unicodedata.name(char))      # BENGALI LETTER KA
print(unicodedata.category(char))  # Lo (Letter, other)

# Check if a character is Bengali
def is_bengali(ch: str) -> bool:
    return "\u0980" <= ch <= "\u09FF"

# Iterate over a Bengali string — conjuncts are multiple code points
text = "বাংলা"  # "Bangla"
for i, ch in enumerate(text):
    print(f"  [{i}] U+{ord(ch):04X} {unicodedata.name(ch, '?')}")

JavaScript

// Regex for Bengali block
const bengaliPattern = /[\u0980-\u09FF]/;

function containsBengali(text) {
  return bengaliPattern.test(text);
}

// Grapheme clusters — important for Bengali
// "কি" is 2 code points but 1 visual unit
const segmenter = new Intl.Segmenter("bn", { granularity: "grapheme" });
const segments = [...segmenter.segment("বাংলা")];
console.log(segments.length); // Visual grapheme count

Sorting Bengali Text

Bengali sorting follows the traditional script order (vowels first, then consonants in systematic phonological order). ICU provides a Bengali-aware collator:

import icu

collator = icu.Collator.createInstance(icu.Locale("bn_BD"))
words = ["বাংলাদেশ", "আমার", "সোনার"]
sorted_words = sorted(words, key=collator.getSortKey)
print(sorted_words)  # Bengali dictionary order

Bengali vs. Assamese

Assamese uses the same script with minor differences:

The shared Unicode block means Bengali and Assamese text are encoded identically at the character level, with the distinction handled by font selection and locale settings.

Key Takeaways

• Bengali script is an abugida used by 300+ million people for Bengali and Assamese, encoded in the Unicode Bengali block (U+0980–U+09FF, 96 characters).
• Conjunct consonants (juktakkhor) — ligatures of 2+ consonants — are the script's most complex feature, requiring extensive OpenType GSUB tables in fonts.
• Vowel signs can appear left, right, above, below, or split around the consonant, and the rendering engine must reorder left-position vowels from logical to visual order.
• The hasanta (virama, U+09CD) is the key combining character — it suppresses the inherent vowel and triggers conjunct formation between consonants.
• Reph (র above), ya-phala (য below), and ra-phala (র below) are special combining forms that appear throughout Bengali text.
• Use quality fonts with full OpenType Bengali support (Noto Sans Bengali, SolaimanLipi) and modern rendering engines (HarfBuzz) to ensure correct display.