The Sound of Softening

If you read a page of Prakrit and then a page of Sanskrit side by side, one thing jumps out immediately: Prakrit words seem to have holes in them. Where Sanskrit has लोकloka, Prakrit has लोअloa. Where Sanskrit has नगरnagara, Prakrit has णयरṇayara. Where Sanskrit has कविkavi, Prakrit has कइkaï.

The consonants have not been forgotten. They have been dropped — deliberately, by rule. This dropping is called lenition, and it is the most characteristic feature of Prakrit phonology. This post is about how it works.

The great rule: nine consonants that vanish

Hemacandra states the rule in one line, but it is the single most important rule in the entire grammar. Between two vowels, when they stand alone (not in a cluster) and not at the very start of a word, these nine consonants are elided:

क, ग, च, ज, त, द, प, य, वk, g, c, j, t, d, p, y, v → nothing

That is one rule. Nine consonants. Every time one of them sits between vowels in the middle of a Sanskrit word, Prakrit lets it drop.

The little y that keeps things smooth

When the elision produces two a vowels next to each other, a small glide ― y sound gets inserted between them. Hemacandra calls this the yaśruti — literally, "a heard y."

So तीर्थकरtīrthakara "one who makes a ford" does not become titthaara but तित्थयरtitthayara. The elided k between the two a's leaves behind a tiny y to bridge them. Similarly नगरnagara becomes णयरṇayara (a soft y filling the gap left by g), and प्रजापतिprajāpati becomes पयावइpayāvaï.

This is not a separate transformation — it is the natural consequence of the elision rule meeting two a vowels. If you see a y in a Prakrit word where the Sanskrit had k, g, c or j, it is almost certainly this glide, not the original y.

Aspirates go to h

Sanskrit's aspirated consonants — the sounds spelled with an हh like खkh, घgh, थth, धdh, भbh — do not vanish. Instead, they all collapse into a single sound: हh.

This is Prakrit's second great sound-shift. Together, the two rules — nine consonants dropping and five aspirates becoming हh — explain the majority of what looks different about Prakrit at first glance.

Some exceptions worth knowing

Not everything falls into line. Two exceptions in particular are common enough that you will meet them constantly.

Word-initial यy becomes जj. The consonant यy only drops when it is between vowels. When it stands at the very start of a word, it hardens into जj instead.

So अवयवavayava (middle y, unchanged) keeps its y, but यशस्yaśas (initial y) shifts to जसjasa. This is why so many Prakrit words in dictionaries start with j- where you might have expected y-.

Between vowels, पp becomes वv, not nothing. The elision rule says पp drops, but a separate rule saves it and turns it into वv instead when it stands between an अa or आā and another vowel.

If you see a v in a Prakrit word where the Sanskrit had p, this is almost always what has happened.

Reading a Prakrit word backwards

If you know these rules, you can now do something useful: given a Prakrit word, guess what Sanskrit word it came from.

Take साहाsāhā. Two vowels, an h, another vowel. The h between vowels in Prakrit almost always comes from one of the five Sanskrit aspirates. Which one? शाखाśākhā, "branch," fits perfectly.

Take जसjasa. Initial j — likely from initial y. Middle s — Prakrit only has one sibilant, so this could be from Sanskrit ś, ṣ, or s. Answer: यशस्yaśas, "fame."

Take मुहmuha. Word-internal h between vowels, again an aspirate. Answer: मुखmukha, "face."

This kind of reverse reading is exactly what Hemacandra's grammar makes possible — a systematic way to walk between the two languages. In the next post we will look at the smaller consonant shifts that give Prakrit its distinctive middle.