तद्धितप्रत्ययाः — the affixes that make words from words

One line separates them from कृत्

Orientation

कृत् attaches to a धातु. तद्धित attaches to a प्रातिपदिक — a noun stem, itself usually already a finished word. भू + क्त्वा gives भूत्वा, "having been" — built on a root. गुरु + त्व gives गुरुत्वम्, "heaviness" — built on गुरु, a word that already means something and already declines on its own.

तद्धिताःThe single sūtra that opens the whole section

— 4.1.76. It sits inside the same अधिकार chain page 16/17 already met — प्रत्ययः 3.1.1, परश्च 3.1.2, आद्युदात्तश्च 3.1.3, ङ्याप्प्रातिपदिकात् 4.1.1 — so every affix from here through the end of the fifth adhyāya is, by definition, a तद्धित: it comes after a ङ्यन्त/आबन्त/प्रातिपदिक word, not after a धातु. That one word, तद्धिताः, is the entire boundary between this page and the last three.

What "प्रातिपदिक" already includes. Not just simple stems like गुरु or धन — a कृदन्त (गच्छत्, कृत — pages 16–18) or even a समास (page 09) is also a प्रातिपदिक, and can itself take a तद्धित affix. The two chapters build on each other, not just sit side by side.

अपत्य — descent, the largest and most characteristic group

4.1.92

"X's son/daughter/descendant" is asked for constantly — in epic and purāṇic Sanskrit almost every person is introduced by patronymic before they are introduced by name. This is also the group where वृद्धि, a term this series has used since the sandhi rules of page 03, finally does visible work of its own.

अण्तस्यापत्यम् — "his offspring"

The sense, and the default affix

— 4.1.92. षष्ठ्यन्त — "of that" — names a person; the affix that follows means "the offspring [of that person]". Which affix depends on the base, but the single most common answer, covering a huge range of names, is अण्, licensed by — 4.1.83 , "before [we reach the sūtra containing the word] दीव्यति, [the affix is] अण्" — i.e. अण् is the default for a long run of the अपत्य section, until something more specific overrides it.

Four instances

Every pair below was checked against the actual grammar engine — every one of the 175 taddhita affixes tried against each base, over CDP, keeping whichever one produced the textbook word (see the .src note further down). Watch the first vowel of each base, and only the first.

Watch one happen, step by step

Pick any of the four. Every step is a real rule the engine actually applied, in the order it applied them.

The exception: अण् is not the only apatya-affix

— 4.1.85 . This is why दैत्य is not *दैत: दिति, अदिति, आदित्य and a handful of पति-final compounds are named outright in this sūtra, and take ण्य instead of the default अण्. The Kāśikā on it gives the same three examples this page uses: दैत्यः, आदित्यः, आदित्यम्. This is the ordinary shape of a Pāṇinian domain — a broad default, and a short, closed list of exceptions stated by name where the default would give the wrong word.

मतुप् / इनि — possession

5.2.94

"Having X" — the second-largest everyday group. A person with wealth is धनवान्; a person with intelligence is बुद्धिमान्. Every word below is one of the corpus's own 9,007 declined stems — none of these six is typed from memory.

मतुप्तदस्यास्त्यस्मिन्निति मतुप्

The rule

— 5.2.94. "That exists in/belongs to this" — मतुप् is the default affix for exactly that sense, added to almost any base.

The alternation: मत् after most sounds, वत् after अ

— 8.2.9 . मतुप्'s own म् turns to व् when the base's last sound is अ (with a short list of exceptions). बुद्धि ends in इ, so मतुप् stays मत्: बुद्धिमान्. बल and भग both end in अ, so it surfaces as वत्: बलवान्, भगवान्.

The alternative after अ: इनि

— 5.2.115 . After a base ending in अ, इनि (or ठन्) may replace मतुप् altogether — अन्यतरस्याम्, "optionally", inherited from the sūtras around it. धन + इनि gives धनी where धन + मतुप् would have given धनवान्; both are real words, and both mean "wealthy man" — the choice is stylistic, not semantic.

त्व / तल् — the affix that lets Sanskrit do philosophy

5.1.119

Turn any quality-word into the abstract noun of that quality — "heaviness", "humanity", "brahman-ness" — and a whole vocabulary of Sanskrit philosophy opens up: सत्यत्वम्, आत्मत्वम्, ब्रह्मत्वम्. One sūtra licenses two different-looking affixes at once, and both are still alive in the language today.

त्व · तल्तस्य भावस्त्वतलौ

— 5.1.119. "The state ( भाव) of that" — एक एव सूत्रम्, two affixes, both meaning exactly the same thing. गुरु is one of the corpus's own headwords; the two abstracts built on it were each read off the live derivation engine, run exactly like the अपत्य words above (see the .src note).

Neither form outranks the other — they are simply two spellings a Sanskrit writer chooses between by ear, the way English keeps both -ness and -ity. Neither triggers वृद्धि: neither त् nor तल् carries a ञ्/ण् इत् marker, so 7.2.117 (below) never fires for this pair — गुरुत्व and गुरुता both keep गुरु's own उ untouched.

तरप् / तमप् — comparing two, comparing many

5.3.55 · 5.3.57

Sanskrit keeps the two-vs-many distinction English politely blurs ("heavier" serves for both "of the two" and "of the many" in casual English; Sanskrit will not let you.)

तरप् · तमप्two different sūtras, two different counts

— 5.3.55 — licenses तमप् in the sense अतिशायन ("excess"), without restriction: the superlative, "-est", among any number.

— 5.3.57. The very next rule narrows the same sense to exactly two (द्विवचनविभज्योपपदे) and licenses तरप् for it instead: the comparative, "-er", strictly between two.

मय — what a thing is made of

4.3.143

A short, everyday group: attach मयट् to a material and get an adjective for anything made of it.

मयट्मयड्वैतयोर्भाषायामभक्ष्याच्छादनयोः

— 4.3.143. In everyday speech (भाषायाम्), for something not eaten ( अभक्ष्य) — covering material, essentially — मयट् (or वति) names what a thing is made of. दारु and हिरण्य are both corpus neuter stems; the two -मय words were each read off the engine. Their own trigger sūtras differ — दारु is on the explicit list at (नित्यं वृद्धशरादिभ्यः, "always, after वृद्ध, शर and the rest of that list"); हिरण्य's मयट् is licensed by a same-affix rule in a different pāda, — but the affix, and what it means, is identical either way.

वृद्धि is the तद्धित signature

7.2.117

Every अपत्य word above changed its first vowel and nothing else. None of त्व, तल्, तरप्, तमप् or मयट् changed anything at all. That is not a coincidence, and once the reason is visible it becomes a reading shortcut.

ञ्णित्Why some taddhita affixes strengthen the base and most do not

The rule

— 7.2.115 — vowels of the अङ्ग take वृद्धि before an affix that is ञित् or णित् (carries an इत् letter ञ् or ण्, front or back). — 7.2.117 — specializes this for तद्धित affixes specifically: it is the base's first vowel, not the one nearest the affix, that strengthens.

Sort the affixes on this page by that one property

प्रत्ययः	इत् letter	वृद्धि?	seen above

The heuristic. Reading unfamiliar Sanskrit, a word whose first vowel is unexpectedly a वृद्धि vowel (आ, ऐ or औ) where a shorter, more familiar word would have had अ/आ, इ/ई or उ/ऊ is very often a तद्धित derivative — usually अपत्य, sometimes another ञ्णित् sense this page has not covered. Strip the वृद्धि back down (औ→उ, ऐ→इ, आ→अ) and what is left is often a real, plainer word: कौरव → कुरु, दैत्य → दिति, वासुदेव → वसुदेव, मानव → मनु. That one move — un-strengthen the first vowel — unlocks a surprising amount of vocabulary without ever opening a dictionary.

The वृद्धि dial

Interactive

Four bases, one operation. Step through them and watch it is always the same move on the same position — never a new rule to memorise per word.

+ →

All 175, as the engine knows them

Coverage

प्रत्ययः	name	adds

Where every form on this page comes from. The 175-row table above is the corpus's own taddhita list, core/vidyut.json, read through C.krt()['taddhita'] — nothing in it is retyped. Every sūtra citation resolves through C.sutra(), which raises the build on a wrong number. Six derived words — बुद्धिमान्, बलवान्, भगवान्, धनी, ज्ञानी, स्वामी — and their bases are corpus paradigms, C.paradigm() pulls, none typed. The remaining derived words — कौरव, मानव, दैत्य, वासुदेव, गुरुत्व, गुरुता, गुरुतर, गुरुतम, लघुतर, लघुतम, दारुमय, हिरण्यमय — are not among the corpus's 9,007 declined headwords (only मानव, दैत्य and वासुदेव happen to be; कौरव is not, and neither is either base word कुरु or दिति). All twelve were read once, directly, off the actual vidyut-prakriya WASM engine that ../Grammar/site/shabdarupa.html boots — its own "तद्धितान्ताः" card tries every one of the 175 affixes against a stem and keeps what survives; this page's build drove that same call over CDP for every base above, and the rule-history each derivation returned is what the step-by-step player under अपत्य encodes. Trying all 175 against गुरु alone, produce a mechanically well-formed word — most of them dictionary curiosities no one actually says, because Pāṇini licenses a taddhita affix by sense, not by shape, and the engine (asked only "does this affix fit this stem") has no way to know which senses are idiomatic.

Check yourself

अभ्यासः

Six questions. Every answer names the rule that decides it.