Historic Indian Scripts (Sharada, Khojki, Nandinagari …)

All eight scripts are descendants of Brahmi, the script of Ashoka’s edicts of the 3rd century BC, and all work the same way: each consonant letter carries a built-in a, a vowel sign changes it (𑆑 is Sharada ka, 𑆑𑆵 kī, 𑆑𑆾 ko), a virama silences it, and the letters run left to right in the varga rows that Sanskrit grammarians fixed over two thousand years ago: ka kha ga gha ṅa, ca cha ja jha ña, and so on. That is why the grid above looks alike from script to script even though the shapes differ completely. Sharada 𑆃, Nandinagari 𑦠, Bhaiksuki 𑰀, Khojki 𑈀, Khudawadi 𑊰, Multani 𑊀 and Dives Akuru 𑤀 are all the vowel a.

They fall into three families. Sharada is the medieval script of Kashmir and the ancestor of the clipped merchants’ scripts of the Indus plain, the Landa scripts, of which Khojki, Khudawadi, Multani and Mahajani are the four in Unicode. Nandinagari and Bhaiksuki are learned scripts for Sanskrit, one from the Hindu south and one from Buddhist monasteries in the east. Dives Akuru is the island script of the Maldives, a southern cousin of the Sinhala and Grantha letters. The Vedic accents and the Devanagari Extended marks are not a script but the pitch and melody marks written into Devanagari when the Vedas are printed for chanting.

Several related scripts have pages of their own, where their whole alphabet sits beside the language that used it: Takri in the Kangri Letters & Script collection, Dogra (Dogri Akkhar) in Dogri Letters & Script, Tirhuta (Mithilakshar) in Maithili Alphabet & Letters, Kaithi in Bhojpuri Alphabet & Letters and Magahi Alphabet & Letters, Tulu-Tigalari in Tulu Script & Letters, Syloti Nagri in Sylheti Alphabet. Brahmi & Kharoshthi, Grantha Script, Modi Script and Siddham Script are separate pages too, and the living Sindhi, Kashmiri and Dhivehi alphabets are in the Sindhi Alphabet, Kashmiri Alphabet & Letters and Dhivehi Letters (Thaana) collections.

Sharada script: the old alphabet of Kashmir

Sharada (𑆯𑆳𑆫𑆢𑆳, Śāradā) is named after the goddess Śāradā, the Kashmiri form of Sarasvati, whose ruined shrine of Sharada Peeth stands in the Neelum valley; Kashmir itself was called Śāradā-deśa, the land of Sharada. The script grew out of the late Gupta letters around the 8th century and was the script of Kashmir, and of much of the western Himalaya and Punjab, for about a thousand years, for Sanskrit above all and for Kashmiri. The great Kashmiri texts, from Kalhana’s Rājataraṅgiṇī to the works of Abhinavagupta and Kashmir Shaivism, were copied in it on birch bark.

Persian and then Urdu displaced it in administration, and in the 20th century Devanagari and the Perso-Arabic alphabet replaced it for Kashmiri, so Sharada survives mainly among Kashmiri Pandits: priests still write horoscopes (the zaatuk or janam-patri) in it, Sharada letters appear on wedding invitations, and a revival movement teaches it online. Its own numerals 𑇐𑇑𑇒𑇓𑇔𑇕𑇖𑇗𑇘𑇙 and the om 𑇄 are in the grid; 𑇄 𑆤𑆩𑆂 𑆯𑆴𑆮𑆳𑆪 is the Shaiva mantra om namaḥ śivāya, and 𑆑𑆳𑆯𑇀𑆩𑆵𑆫 spells Kāśmīra.

Unicode added Sharada in version 6.1 (2012), with extra manuscript marks in 7.0, 8.0 and 13.0. Among the letters, note the Kashmiri vowel signs: the prishthamatra e 𑇎 is written to the left of its consonant, as in old manuscripts, and the inverted candrabindu 𑇏 marks a nasal vowel. Sharada’s daughters are the Landa scripts below and, through them, Gurmukhi, the script of Punjabi; Takri of Himachal and Jammu is its closest relative and sits in the Kangri Letters & Script collection. The modern Kashmiri alphabet, written in Perso-Arabic letters, is in the Kashmiri Alphabet & Letters collection.

Khojki, Khudabadi, Multani and Mahajani: the merchants’ scripts of the Indus

Landa means ‘clipped’ or ‘tailless’: these are the quick, stripped-down scripts that shopkeepers, bankers and traders of Sindh, Punjab, Rajasthan and the Multan region used for letters and account books, often with no vowel signs at all, because a merchant who knew his customers’ names did not need them. Every town had its own variety, most never printed. Unicode has four of them. Khojki (𑈉𑈲𑈐𑈵𑈈𑈮) belongs to the Khoja community of Sindh and Gujarat, Ismaili Muslims who wrote their ginans, devotional hymns attributed to pirs such as Pir Sadardin, in it from about the 16th century, and printed them lithographically in Bombay from the late 19th century. It is the most complete of the four, with vowel signs, the dandas 𑈸 𑈹, a word separator 𑈺, the Arabic-style shadda 𑈷 and sukun 𑈾 for Arabic and Persian words, and the doubled letters 𑈋 𑈑 𑈜 𑈢 for Sindhi’s implosive consonants.

Khudawadi, usually spelt Khudabadi after the town of Khudabad on the Indus, was the script Sindhi Hindu traders kept their records in. In 1868 the Bombay Presidency had Narayan Jagannath Vaidya, its deputy educational inspector in Sindh, standardise it, with a full set of vowels, as the script of the Hindu Sindhi schools, which is why its grid is complete: ten vowels, nine vowel signs, the four implosives 𑊽 𑋃 𑋉 𑋕, the flap 𑋊 and its own digits 𑋰𑋱𑋲𑋳𑋴𑋵𑋶𑋷𑋸𑋹. 𑋝𑋡𑋑𑋪𑋐𑋢 spells Sindhī. After 1947 most Sindhi Hindus moved to India and the script fell out of everyday use; a few families still keep accounts in it. The alphabet Sindhi is written in today is in the Sindhi Alphabet and Sindhi Devanagari Alphabet collections.

Multani (𑊠𑊂𑊣𑊖𑊀𑊚, Multān) is the Landa script of Multan and the Saraiki-speaking south of Punjab. It has four vowel letters 𑊀 𑊁 𑊂 𑊃, no vowel signs and no digits, and one punctuation mark, 𑊩; it served everyday letters and trade, and the first book ever printed in it was a New Testament from the Serampore mission press in 1819. The modern Saraiki alphabet is in the Saraiki Alphabet collection. Mahajani is the mahājan’s, the banker’s, script of Rajasthan, Haryana, Punjab and the Delhi region, used for Marwari, Hindi and Punjabi account books (bahi) and taught in merchant schools as part of a trader’s education. It too has only vowel letters 𑅐 𑅑 𑅒 𑅓 𑅔, whose use was optional, and no vowel signs or digits of its own; its one ligature is 𑅶 śrī (𑅶 𑅬𑅱𑅛𑅧, śrī mahājan), the auspicious word at the head of letters and ledgers. Khojki, Khudawadi and Mahajani entered Unicode in version 7.0 (2014), Multani in 8.0 (2015).

Nandinagari, Bhaiksuki and Dives Akuru: scripts of scholars, monks and islanders

Nandinagari (𑧁𑧁𑧠𑦿𑧒𑧁𑧑𑦰𑧈𑧓) is the southern sister of Devanagari: a Nagari script used from about the 8th to the 19th century in Karnataka, Andhra and the Tamil country for Sanskrit manuscripts, inscriptions and the copper-plate grants of the Vijayanagara kings. The followers of Madhva, the Dvaita philosopher of Udupi, copied their Sanskrit texts in it, and it is on their manuscripts that much of the surviving material rests. The letters hang from a headstroke like Devanagari, but with rounder southern shapes; the e sign 𑧤 is written before its consonant, and texts begin with the siddham sign 𑧢 (𑧢 𑧋𑧠𑧈𑧓, siddham śrī). Nandinagari has no digits of its own; scribes used Kannada or Telugu numerals. Unicode 12.0 (2019).

Bhaiksuki (𑰤𑰲𑰟𑰿𑰠 𑰠𑰨𑰿𑰦 𑰭𑰽𑰑, Buddha, Dharma, Saṅgha) is the ‘arrow-headed script’, so called because every letter is topped with a small arrowhead; its Sanskrit name means ‘of the monks’, and older scholarship called it Sindhura. It was used in the 11th and 12th centuries in Bihar and Bengal, in the Pala-era monasteries of the Buddhist east, and only four manuscripts and eleven inscriptions in it are known, among them the Candrālaṃkāra, the Maṇicūḍajātaka and the Abhidharmasamuccayakārikā. Its number system is unusual: besides digits 𑱑𑱒𑱓𑱔𑱕𑱖𑱗𑱘𑱙𑱐 it has separate signs for one to nine 𑱚 𑱛 𑱜, for the tens 𑱣 𑱤 and a hundreds mark 𑱬, the additive way Indian manuscripts numbered their leaves. Unicode 9.0 (2016).

Dives Akuru (𑤝𑤱𑤩𑤵𑤭𑤱 𑤀𑤌𑤳𑤧𑤳, Dhivehi Akuru, ‘island letters’) is the script the Maldives wrote Dhivehi in before Thaana: the later form of Eveyla Akuru, used from about the 14th century on coral-stone inscriptions, copper-plate grants (lōmāfānu) and paper. Thaana, written right to left, appeared in the 18th century and gradually replaced it, but Dives Akuru stayed in use beside Thaana in the southern atolls to the end of the 19th century, the last known official document in it was written in 1927, and H. C. P. Bell, who named its older form Evēla Akuru, found an astrology book in it on Addu atoll. It reads left to right, has its own digits 𑥐𑥑𑥒𑥓𑥔𑥕𑥖𑥗𑥘𑥙, a halanta 𑤽 and virama 𑤾, the prefixed nasal 𑤿 that writes a nasal before a consonant, and the joined forms medial ya 𑥀, initial ra 𑥁 and medial ra 𑥂. Unicode 13.0 (2020). Today’s Dhivehi alphabet is in the Dhivehi Letters (Thaana) collection.

Vedic accents and Samaveda marks, and how to make all of this display

Vedic Sanskrit had a pitch accent that classical Sanskrit lost, and reciters still chant it. Printed Rigveda texts mark it with two signs from the Devanagari block: the underline ॒ (anudātta, the low syllable) and the vertical stroke ॑ (svarita, the falling syllable), the high udātta being left unmarked, so the first words of the Rigveda read अ॒ग्निमी॑ळे पु॒रोहि॑तं. ᳚ and ᳛ mark a doubled or tripled svarita. Yajurveda schools (the Maitrāyaṇī, Kāṭhaka and Taittirīya traditions) and the Atharvaveda use their own signs for the independent svarita, such as ᳕ ᳖ ᳗ ᳡, and the Samaveda, which is sung, writes small digits above the syllables for the notes of the melody: ꣡ ꣢ ꣣ ꣤, with the small letters ꣪ ꣫ ꣬ ꣯ as further singing instructions. These are the Vedic Extensions block, added in Unicode 5.2 (2009) from proposals worked out with Vedic scholars, and the Devanagari Extended block of the same version.

The same blocks hold the signs of Sanskrit manuscripts rather than of chant: the candrabindu-virama ꣳ and its double ꣴ, which write the Vedic nasal gum, the flower-like pushpika ꣸ used as a divider (॥ ꣸ ॥), the gap filler ꣹ that a scribe drew across a damaged patch, the caret ꣺ for an insertion, the siddham sign ꣼ that opens a text (꣼ ॥ श्रीगणेशाय नमः ॥) and the Jain om ꣽ. Unicode 15.0 (2022) added the head marks and bhale signs 𑬀 𑬂 𑬆, the auspicious opening marks written at the head of letters and account documents in Gujarat and Rajasthan, several of them shaped like the figures 5 and 9, which is what their Unicode names describe.

None of these scripts is on an ordinary keyboard, and most phones and computers have no font for them: a Sharada or Khojki letter shows as an empty box unless a font that has it is installed. Google’s free Noto fonts cover every block here: Noto Sans Sharada, Noto Sans Khojki, Noto Sans Khudawadi, Noto Sans Multani, Noto Sans Mahajani, Noto Sans Bhaiksuki, Noto Sans Nandinagari and Noto Serif Dives Akuru, and Noto Sans Devanagari for the Vedic and extended marks. Install the one you need, then copy the letters from the grid above: tap a character to copy it. In HTML or CSS you can also write a character by its code, 𑆃 for Sharada 𑆃. The accent marks are combining characters, so paste them after the syllable they belong to, and never substitute look-alikes: the Sharada danda 𑇅 is not the Devanagari ।, and the Sharada om 𑇄 is not ॐ.