Tamil Brahmi Converter

Background

What Tamil-Brahmi Is

Brahmi is the ancient script from which most scripts of South and Southeast Asia descend. When it was first used for Tamil, it had to change: Brahmi was built for Prakrit, and Tamil has sounds Prakrit does not. Tamil writers added letters for four of them, ழ, ள, ற and ன, adapted from the Brahmi letters with the nearest sounds. They also needed a way to show a consonant with no vowel, which Tamil words end in all the time and Prakrit words rarely do.

The standard study of these inscriptions is Iravatham Mahadevan’s Early Tamil Epigraphy: From the Earliest Times to the Sixth Century A.D. (Cre-A, Chennai, and the Department of Sanskrit and Indian Studies, Harvard University, Harvard Oriental Series 62, 2003). It edits the Tamil-Brahmi and early Vatteluttu inscriptions from about the 2nd century BCE to the 6th century CE, cut in caves or found on pottery, coins, seals and rings, with letter charts for the script. Mahadevan placed the earliest of them around the end of the 3rd century or the start of the 2nd century BCE, judging by the shapes of the letters.

Excavations since then have reopened the question of how early the script began. In 2019 the Tamil Nadu State Department of Archaeology published Keeladi: An Urban Settlement of Sangam Age on the Banks of River Vaigai, on the excavations at Keezhadi (Keeladi) near Madurai. One of six samples from the fourth season, taken at a depth of 353 cm, was dated by accelerator mass spectrometry to 580 BCE, and the department said this pushed Tamil-Brahmi back to the 6th century BCE. Other archaeologists have questioned whether a date from that sample also dates the inscribed potsherds. The earliest date of Tamil-Brahmi is still debated.

How Unicode Encodes It

Unicode has one Brahmi block, U+11000 to U+1107F, shared by every form of the script. The three Tamil consonants ழ, ற and ன were part of it from the start, as BRAHMI LETTER OLD TAMIL LLLA, RRA and NNNA (U+11035 to U+11037). Unicode 14.0, in 2021, added six more characters for Old Tamil, proposed by Vinodh Rajan and Shriramana Sharma: the dot-shaped Old Tamil virama, short e and short o as letters and as vowel signs, and the Tamil-Brahmi form of ள (U+11070 to U+11075). This converter uses all of them, so Tamil’s short and long vowels stay distinct and nothing is lost on the way.

Dead consonants

Pulli Dot, Virama or Not Marked

In modern Tamil, the pulli turns க (ka) into க் (k). Early Tamil-Brahmi did not always mark this. Mahadevan described three stages. In the first, a plain consonant had no built-in vowel, so it read as k, and one stroke marked a following a or ā. In the second, the stroke marked only long ā, so a plain letter could mean k or ka and the reader had to decide. In the third, the pulli dot marked the vowel-less consonant, and the writing became unambiguous. Some scholars have since questioned how these stages are dated, but the ambiguity they describe is why a modern rendering has to choose.

• Pulli dot (default) uses BRAHMI SIGN OLD TAMIL VIRAMA, U+11070. It matches the third stage and modern Tamil, never joins letters into conjuncts, and can be converted back to Tamil without guessing.
• Virama uses the generic BRAHMI VIRAMA, U+11046, for fonts that do not have the Old Tamil sign. It is drawn as a stroke rather than a dot, and some fonts join it with the next consonant into a conjunct, which Tamil-Brahmi did not do. That is why the pulli dot is the default.
• Not marked drops the pulli, as many early inscriptions did. வணக்கம் becomes 𑀯𑀡𑀓𑀓𑀫, where k and ka look the same.

An honest note

What This Converter Does and Does Not Do

It writes modern Tamil in Tamil-Brahmi letters, one letter for one letter. It does not translate into Old Tamil, change spelling to match the inscriptions, or imitate the letter shapes of a particular site or century. The shapes you see come from the Noto Sans Brahmi font, which draws a standard form of each letter.

• Grantha letters (ஜ ஷ ஸ ஹ, and ஶ) map to the Brahmi letters for the same sounds. This is a modern convenience: it shows the sound, not a spelling found in the inscriptions.
• The aytham ஃ has no Unicode Brahmi character and is kept as it is.
• Numbers are kept as typed, including Tamil numerals such as ௧ ௨ ௩.
• Punctuation, spaces and English words you leave in pass through unchanged. Tanglish is converted to Tamil first by our typing engine, which can guess a word wrong; check the line under your text.