10: Grammar: Derived words (`taddhita`, `uṇādi`)

Info

Part of “A Beginner (Non-Linguist) Guide to Learning Pali”

How this chapter fits

Chapter 1 introduced the affixes in its sections on taddhita, uṇādi and sādhana; this chapter deepens them, with the senses from my translation of Kaccāyana’s Taddhitakappa and Uṇādikappa. Its examples and exercises are words from the texts of chapters 3, 5, 7 and 9 and the bonus chapter. The text that follows it is the bonus chapter, the core teachings.

Besides the compounds of chapter 6, Pāḷi makes new nouns by adding an affix (paccaya) to a word that already exists. A taddhita affix is added to a noun: vāsiṭṭho (“Vasiṭṭha’s descendant”) is vasiṭṭha with ṇa added and its first vowel lengthened {344}. An uṇādi affix is added to a root: samatho (“calm”) is √samu with tha {628}. The kita affixes, which make the participles, absolutives and infinitives marked 🔽 in the texts, are the subject of 8.8 and are tabled in Appendix J.

flowchart LR
  W["a derived word"] --> T["from a noun: taddhita §344–§405"]
  W --> K["from a root: kita §524–§623, uṇādi §624–§673"]
  T --> T1["belonging, descent: ṇa, ṇika"]
  T --> T2["having: vantu, mantu, ī"]
  T --> T3["the abstract state: tā, tta"]
  T --> T4["numbers and manner: ma, tiya, dhā, so"]
  K --> K1["the doer: aka, tu"]
  K --> K2["the act: ana, ti"]
  K --> K3["to be done: tabba, anīya, ya"]

vasiṭṭhassa apaccaṁ → vāsiṭṭho

🚹⨀⑥(vasiṭṭhassa) 🚻⨀①(apaccaṁ) → 🚹⨀①🔽(ṇa)(vāsiṭṭho)
of Vasiṭṭha / descendant → Vasiṭṭha’s descendant
The phrase becomes one word: the vibhatti ending of vasiṭṭha and the word apacca drop away, the affix takes their place, and the new word takes its own ending.

The derived word is marked 🔽 in this book, with the affix as Kaccāyana names it: 🚹⨀①🔽(ṇa)(vāsiṭṭho), 🚹⨀①🔽(tha)(samatho). Many affixes are written with a ṇ that is not part of the affix: it is a signal that the first vowel of the base is lengthened, a to ā, i to e, u to o {400}, and the ṇ itself is dropped {396}. The affix ṇa therefore adds only a; ṇika adds ika.

The affixes you will meet most

Kaccāyana treats some sixty taddhita affixes and fifty uṇādi affixes. A handful account for most of the derived words in the texts. The counts are of the words the Digital Pāḷi Dictionary assigns to each affix and of their occurrences in the early discourses¹:

affix	sense	a word you will meet	words	occurrences	section
vantu, mantu	having	bhagavā (“the Blessed One”), āyasmā (“venerable”)	222	45,000	10.5
ka	a small or ordinary one of its kind, an adjective	pāpako (“bad”)	1,300	19,000	10.3
ṇa	a descendant, belonging to	vāsiṭṭho, māṇavo (“a young brahmin”)	685	64,000	10.1
ṇika, ika	connected with, living by	nāviko (“a sailor”), santiko (“near”)	513	7,600	10.2
tā, tta	the abstract state, “-ness”	devatā (“a deity”), sahabyatā (“companionship”)	1,108	10,200	10.4
tara, tama	comparison, “more”, “most”	aññataro (“a certain one”), katamo (“which?”)	183	17,600	10.4
tiya, ma	the ordinals	dutiyo (“second”), pañcamo (“fifth”)	33	12,900	10.6
dhā, so, thā, thaṁ	division and manner	sabbaso (“in every way”), yathā (“as”), kathaṁ (“how?”)	82	21,200	10.6
aka, tu (uṇādi)	the doer	kārako (“a doer”), satthā (“the teacher”)	368	16,800	10.7
ana, ti (uṇādi)	the act	ṭhānaṁ (“a place”), āpatti (“an offence”)	1,697	87,000	10.7

Kaccāyana orders the taddhita affixes by the sense they carry, from the descendant {344} to the ordinals and the adverbs {373}–{399}, and ends with the changes the affixes bring to the base {400}–{405}. The uṇādi rules first say what the affixes denote, the agent, the action, the purpose, and then list the affixes that form individual words. The chapter follows that order. Appendix H tables every taddhita rule and Appendix K every uṇādi rule.

section	sense	affixes	rules
10.1	the descendant, and the shape of an affixed word	ṇa, ṇāyana, ṇāna, ṇeyya, ṇi, ṇava, ṇera; what the ṇ signals	{344}–{349}, {396}, {400}–{405}
10.2	mixed with, crossing by, studying, made by, and the other senses	ṇika, ṇa, ima, iya	{350}–{353}
10.3	a collection, an occasion, a likeness, a dependence, an abundance	kaṇ, ṇa, tā, iya, āyitatta, la, ālu	{354}–{359}
10.4	the abstract state, and comparison	ṇya, tta, tā, ttana, ṇa, kaṇ; tara, tama, isika, iya, iṭṭha	{360}–{363}
10.5	“having that”, and “made of that”	vī, sī, ika, ī, ra, vantu, mantu, ṇa; maya	{364}–{372}
10.6	the numerals: ordinals, the forms of the numbers, division and manner	ma, tha, ṭha, tiya, ī, ka; dhā, so, thā, thaṁ	{373}–{399}
10.7	uṇādi: the agent, the action and the purpose	kita, kicca, tta, yu, ṇvu, ṇī, tuṁ and the rest	{624}–{626}, {635}–{637}, {639}–{642}, {650}–{655}
10.8	uṇādi: the affixes that form individual words	tha, ma, la, ya, tti, īvara, ūra, ṭha and the rest	{627}–{634}, {638}, {643}–{649}, {656}–{673}

Reading path

On a first reading take 10.1 for the shape of an affixed word, then 10.5, 10.4 and 10.7, the affixes of the table above; the sections on the rarer senses and on the individual uṇādi words are for looking things up, and every rule number links to the full rule in the Kaccāyana translation.

Footnotes

1. The Digital Pāḷi Dictionary is the place to look up what a word means, but it is weak at saying how a derived word is made: its construction and compound-type fields often cut a word differently from Kaccāyana, deriving kaccāyano from kati rather than kacca, madaniyaṁ from √mad with anīya rather than madana with iya, and reading rūpasaññā as a ⑥ compound rather than the ⑦ of Kaccāyana’s rule. Where the two differ, this book follows Kaccāyana: the examples are the rules’ own, and the dictionary’s cut is no evidence against them. ↩