The first section of the Padarūpasiddhi gathers Kaccāyana’s definitions (saññā) and the sandhi rules of the Sandhikappa in Buddhappiya’s own order — vowel sandhi, the cases where the vowel stays unchanged (pakati), consonant sandhi, niggahīta sandhi — and derives every example word it cites, quoting the rules by their wording. Each sutta carries the Kaccāyana number it corresponds to and a card to the Kaccāyana page, where the rule’s own analysis, examples and counter-examples are.
Buddhappiya’s order for the sandhi rules, with the Rūpasiddhi numbers and the Kaccāyana rules each part draws on.
flowchart LR
S(["Sandhikaṇḍa 1–59"]) --> A["Saññāvidhāna 1–11<br/>letters, vowels, consonants, the terms<br/>§1–§9, with §602–§603 on heavy syllables"]
S --> B["Sarasandhividhāna 12–34<br/>a vowel before a vowel: elision 13, the option 15, lengthening 17–18,<br/>y and v for e and o 19–20, insertion 34<br/>§10–§22, §35, §42–§47 and three noun rules"]
S --> C["Pakatisandhividhāna 35–36<br/>the vowel unchanged §23–§24"]
S --> D["Byañjanasandhividhāna 37–48<br/>a vowel before a consonant: long, short or elided 37–39,<br/>doubling 40 and 42, o for ava 45<br/>§25–§29, §36, §48–§50 and three noun rules"]
S --> E["Niggahītasandhividhāna 49–59<br/>ṃ before vowels and consonants, and the closing rule 59<br/>§30–§34, §37–§41, §51"]
Tattha jinasāsanādhigamassa akkharakosallamūlakattā taṃ sampādetabbanti dassetuṃ abhidheyyappayojanavākyamidamuccate.
There [after the opening verses], since the understanding of the Conqueror’s teaching has skill in letters (akkharakosalla) as its root, this sentence, which states the subject-matter (abhidheyya) and the purpose (payojana) [of the grammar], is spoken in order to show that that [skill] is to be acquired.
Buddhappiya keeps Kaccāyana’s first sutta first. It is not an operating rule but the sentence that names the subject of the whole grammar (letters, and the words made of them) and its purpose (the understanding of the Buddha’s word), and he places it before the Saññāvidhāna, the assignment of technical terms that begins at 2 (§2): Akkharāpādayo ekacattālīsaṃ ; the badge sīhagatika marks a rule whose scope is everything that follows. His gloss widens Kaccāyana’s sabbavacanānamattho to “the meaning of every word, worldly (lokiya) or supramundane (lokuttara)”, names the typical fault — an unaspirated letter (sithila) read for an aspirated one (dhanita), terms he will define at 11 (§9): Parasamaññā payoge — and adds that words (pada) themselves count as akkhara, being nothing but collections of letters, so that the sutta covers the whole of grammar and not only the alphabet. The verse then turns the statement into an exhortation: skill in letters is to be acquired from a teacher, with the five duties towards him. There is no option marker.
Yo koci lokiyalokuttarādibhedo vacanattho, so sabbo akkhareheva saññāyate. Sithiladhanitādiakkharavipattiyañhi atthassa dunnayatā hoti, tasmā akkharakosallaṃ bahūpakāraṃ buddhavacanesu, ettha padānipi akkharasannipātarūpattā akkharesveva saṅgayhanti.
Whatever meaning of a word there is, of the kinds worldly (lokiya), supramundane (lokuttara) and the rest — all of it is made known (saññāyate) only through letters (akkhara). For when a letter goes wrong — an unaspirated (sithila) letter for an aspirated (dhanita) one, and the like — the meaning becomes hard to make out; therefore skill in letters is of great help in the words of the Buddha. Here words (pada) too, since in form they are collections of letters, are included under “letters”.
Tasmā akkharakosallaṃ, sampādeyya hitatthiko;
Therefore one who seeks his own good should acquire skill in letters,
Tatthādo tāva saddalakkhaṇe vohāraviññāpanatthaṃ saññāvidhānamārabhīyate.
There, first of all, in order to make known the technical usage (vohāra) of the science of words (saddalakkhaṇa), the assignment of technical terms (saññāvidhāna) is begun.
The first of the ten definitions of the Saññāvidhāna (2–11), which follow Kaccāyana’s §2–§9 in order except that Buddhappiya inserts the two garu rules of the Kibbidhānakappa, 6 (§602): Dumhi garu and 7 (§603): Dīgho ca , after the long vowels. Kaccāyana’s vutti only lists the forty-one letters; Buddhappiya adds everything a beginner needs to say the list: the etymology of akkhara (“they do not wear away”), the principle behind the order of the letters (place, organ and effort of articulation — ṭhāna, karaṇa, payatana), the six places with the letters born in each, the organ and the effort for each class, the reason vowels come before consonants (they are the support, nissaya, the consonants the supported, nissita), and the point of the number: the count of forty-one is a boundary, and a text that uses more letters — the extra vowels and sibilants of Sanskrit — is by that token not the Buddha’s word. His last remark explains the sutta’s api as a back-reference to the letters already spoken of in 1 (§1): Attho akkharasaññāto . There is no option marker.
Akkharā api ādayo ekacattālīsaṃ, te ca kho jinavacanānurūpā akārādayo niggahītantā ekacattālīsamattā vaṇṇā paccekaṃ akkharā nāma honti. Taṃ yathā – a1
The letters (akkhara) too, beginning with a, are forty-one: those forty-one sounds (vaṇṇa), suited to the Conqueror’s word, from a to the niggahīta, are each called akkhara. Namely: a ā i ī u ū e o, ka kha ga gha ṅa, ca cha ja jha ña, ṭa ṭha ḍa ḍha ṇa, ta tha da dha na, pa pha ba bha ma, ya ra la va sa ha ḷa, aṃ — these are the letters.
Nakkharantīti akkharā, a ādi yesaṃ te ādayo. Akārādīnamanukkamo panesa ṭhānādikkamasannissito, tathā hi ṭhānakaraṇappayatanehi vaṇṇā jāyante, tattha cha ṭhānāni kaṇṭhatālumuddhadantaoṭṭhanāsikāvasena.
They are akkhara because they do not wear away (na kharanti); ādayo are those whose beginning is a. This sequence of a and the rest rests on the sequence of place [of articulation] and so on: for sounds are produced by place (ṭhāna), organ (karaṇa) and effort (payatana). Of these, the places are six: throat, palate, head, teeth, lips and nose.
Tattha – a vaṇṇa ka vagga ha kārā kaṇṭhajā.
Of these: the a-vaṇṇa [a and ā], the ka-group and ha are born in the throat.
I vaṇṇa ca vagga ya kārā tālujā. Ṭa vagga ra kāra ḷa kārā muddhajā. Ta vagga la kāra sa kārā dantajā. U vaṇṇa pa vaggā oṭṭhajā. E kāro kaṇṭhatālujo. O kāro kaṇṭhoṭṭhajo.
The i-vaṇṇa, the ca-group and ya are born at the palate. The ṭa-group, ra and ḷa are born at the head [the roof of the mouth]. The ta-group, la and sa are born at the teeth. The u-vaṇṇa and the pa-group are born at the lips. e is born at throat and palate. o is born at throat and lips.
Va kāro dantoṭṭhajo.
va is born at teeth and lips.
Niggahītaṃ nāsikaṭṭhānajaṃ.
The niggahīta is born at the place of the nose.
Ṅa ña ṇa na mā sakaṭṭhānajā, nāsikaṭṭhānajā cāti.
ṅa, ña, ṇa, na and ma are born both at their own place and at the place of the nose.
Ha kāraṃ pañcameheva, antaṭṭhāhi ca saṃyutaṃ;
ha joined with the fifth [letter of a group] or with a semivowel (antaṭṭha) orasa); unjoined, it is throat-born.
Karaṇaṃ jivhāmajjhaṃ tālujānaṃ, jivhopaggaṃ muddhajānaṃ, jivhāggaṃ dantajānaṃ, sesā sakaṭṭhānakaraṇā.
The organ (karaṇa) is the middle of the tongue for those born at the palate, the part near the tip of the tongue for those born at the head, the tip of the tongue for those born at the teeth; the rest have their own place as organ.
Payatanaṃ saṃvutādikaraṇaviseso. Saṃvutattama kārassa, vivaṭattaṃ sesasarānaṃ sa kāraha kārānañca, phuṭṭhaṃ vaggānaṃ, īsaṃphuṭṭhaṃ yaralavānanti.
Effort (payatana) is the particular manner of the organ: closed (saṃvuta) and so on. Closure belongs to a; openness (vivaṭa) to the other vowels and to sa and ha; contact (phuṭṭha) to the group letters; slight contact (īsaṃphuṭṭha) to ya, ra, la and va.
Evaṃ ṭhānakaraṇappayatanasutikālabhinnesu akkharesu sarā nissayā, itare nissitā. Tattha –
Thus, among the letters, which differ in place, organ, effort, sound (suti) and duration (kāla), the vowels are the support (nissaya) and the others the supported (nissita). On this –
Nissayādo sarā vuttā, byañjanā nissitā tato;
The vowels are stated first, as the support, and then the consonants, the supported;
Vuttañca –
And it is said:
‘‘Pañcannaṃ pana ṭhānānaṃ, paṭipāṭivasāpi ca;
“And by way of the sequence of the five places too,
Ekenādhikā cattālīsaṃ ekacattālīsaṃ, etena gaṇanaparicchedena –
Forty exceeded by one is forty-one; by this delimitation of the number –
Adhikakkharavantāni, ekatālīsato ito;
[Texts] that have more letters than these forty-one
Api ggahaṇaṃ heṭṭhā vuttānaṃ apekkhākaraṇatthaṃ.
The inclusion of api (“too”) is for the purpose of referring back to those [letters] spoken of before [in the first sutta].
The second definition, in Kaccāyana’s place: the letters have been listed, and now the first eight of them are named sara. Buddhappiya’s gloss adds what the Kaccāyana vutti leaves implicit — the analysis of odantā as “those whose end is o”, with the note that its d is a sandhi augment (the d of 34 (§35): Yava ma da na ta ra lā cāgamā ), and two etymologies of sara: the vowels “go” (saranti) by themselves, or they “make the consonants go” (sārenti), which is the nissaya–nissita relation of the previous sutta stated as a name. He closes by marking the anuvutti: tattha (“among these”) is carried into the next two definitions. There is no option marker.
Tattha tesu akkharesu a kārādīsu o kārantā aṭṭha akkharā sarā nāma honti. Taṃ yathā – a ā i ī u ū e o-iti sarā.
There, among those letters beginning with a, the eight letters that end with o are called sara (vowels). Namely: a ā i ī u ū e o — these are the vowels.
sarameasure
rassa | dīgha
a
one
short
ā
two
long
i
one
short
ī
two
long
u
one
short
ū
two
long
e
two
long
o
two
long
O anto yesaṃ te odantā, da kāro sandhijo, saranti gacchantīti sarā, byañjane sārentītipi sarā.
Those whose end is o are odantā; the d [in odantā] arises from sandhi. They are sara because they go (saranti), that is, proceed [on their own]; or again sara because they make the consonants proceed (sārenti).
‘‘Tatthā’’ti vattate.
Tattha (“among these”) continues [into the following rules].
Third definition, again in Kaccāyana’s order, with tattha carried from 3 (§3): Tatthodantā sarā aṭṭha so that “among the eight vowels” need not be said. Kaccāyana’s vutti merely names a, i, u; Buddhappiya defines the measure (mattā) by which they are short — the time of a snap of the fingers or a blink of the eye — and states the whole scale at once: short vowels one measure, long vowels two, consonants half a measure (the aḍḍhamattā he will use again at 8 (§6): Sesā byañjanā ). He also explains why the sutta says lahumattā (“of light measure”) rather than ekamattā (“of one measure”): so that in verse a vowel drawn out to one and a half measures still counts as short. Two etymologies of rassa close the gloss, and sara and rassa are carried on into the next rule. There is no option marker.
Tattha aṭṭhasu saresu lahumattā tayo sarā rassā nāma honti. Taṃ yathā – a i u-iti rassā.
There, among the eight vowels, the three vowels of light measure are called rassa (short). Namely: a i u — these are the short vowels.
rassameasure
as in
a
one
🚹①⨀(nara) naro, a man
i
one
🚹①⨀(aggi) aggi, fire
u
one
🚹①⨀(bhikkhu) bhikkhu, a monk
Lahukā mattā pamāṇaṃ yesaṃ te lahumattā, mattā saddo cettha accharāsaṅghātaakkhinimīlanasaṅkhātaṃ kālaṃ vadati, tāya mattāya e kamattā rassā, dvimattā dīghā, aḍḍhamattā byañjanā. Lahu ggahaṇañcettha chandasi diyaḍḍhamattassāpi gahaṇatthaṃ. Rassakālayogato rassā, rassakālavanto vā rassā.
Those whose measure (mattā), their quantity, is light are lahumattā. The word mattā here denotes the time reckoned as a snap of the fingers or a blink of the eye; by that measure the short vowels have one measure, the long ones two measures, the consonants half a measure. And the word lahu (“light”) is used here so that in verse (chandasi) a [vowel] of even one and a half measures may be included [as short]. They are rassa because of their connection with a short time, or rassa because they possess a short time.
Sararassa ggahaṇāni ca vattante.
The words sara and rassa also continue.
The last of the vowel definitions, completing the pair with 4 (§4): Lahumattā tayo rassā , whose sara and rassa it takes over (“other than the short ones”). Buddhappiya explains the choice of aññe (“others”) over dvimattā (“of two measures”) exactly as he explained lahu: a vowel drawn to one and a half measures in verse must count as long as well as short, and only a relative term admits it. Then comes an observation that has no counterpart in Kaccāyana: in some places e and o before a conjunct consonant are pronounced as if short (ettha, seyyo, oṭṭho, sotthi). The kvaci here is Buddhappiya’s own, not the sutta’s, and he uses it as Ruiz-Falqués describes the marker (Boundaries and Domains §4.2): to isolate an exception from the general state of affairs — for in maṃ ce tvaṃ nikhaṇaṃ vane and putto tyāhaṃ mahārāja the same vowels before a conjunct keep their full length, as the metre requires. The sutta itself has no option marker.
Tattha aṭṭhasu saresu rassehi aññe dvimattā pañca sarā dīghā nāma honti. Taṃ yathā – ā ī ū e o-iti dīghā.
There, among the eight vowels, the five vowels other than the short ones, having two measures, are called dīgha (long). Namely: ā ī ū e o — these are the long vowels.
Añña ggahaṇaṃ diyaḍḍhamattikānampi saṅgahaṇatthaṃ. Dīghakāle niyuttā, tabbanto vā dīghā. Kvaci saṃyogapubbā ekārokārā rassā iva vuccante. Yathā – ettha, seyyo, oṭṭho, sotthi.
The word añña (“other”) is used so as to include also [vowels] of one and a half measures. They are dīgha because they are set in a long time, or because they possess it. In some places (kvaci) e and o standing before a conjunct are pronounced as if they were short. For example: ettha, seyyo, oṭṭho, sotthi.
lemma
inflected
meaning
🔼(ettha)
ettha
here — e before tth
🚹①⨀(seyya)
seyyo
better — e before yy
🚹①⨀(oṭṭha)
oṭṭho
a lip; a camel — o before ṭṭh
🚺①⨀(sotthi)
sotthi
safety — o before tth
Kvacī ti kiṃ? Maṃ ce tvaṃ nikhaṇaṃ vane. Putto tyāhaṃ mahārāja.
Why “in some places”? [Because in] maṃ ce tvaṃ nikhaṇaṃ vane and putto tyāhaṃ mahārāja [the e and o before a conjunct remain long].
By saying kvaci (in some places) the shortening of e and o before a conjunct is not general: in these verses the e of ce before tv and the o of putto before ty keep their two measures.
mañce tvaṃ nikhaṇaṃ vane (23J 22.1.1: Mūgapakkhajātaka)
putto tyāhaṃ mahārāja (22J 1.1.7: Kaṭṭhahārijātaka)
The canon writes the first phrase mañce, with the niggahīta of maṃ assimilated to c.
Buddhappiya lifts the two rules on the weight of syllables out of Kaccāyana’s Kibbidhānakappa — §602 and §603 stand there to account for the short vowel of datvā and hutvā — and sets them among the definitions, directly after the long vowels of 5 (§5): Aññe dīghā , because garu and lahu are properties of letters and the sandhi rules that invoke the metre (chandānurakkhaṇa, from 34 (§35): Yava ma da na ta ra lā cāgamā onward) presuppose them. The Bālāvatāra (¶45) follows him and states the whole doctrine in a line. His gloss explains the sutta’s odd word: du is “a group of two”, so dumhi means “before a conjunct letter” (saṃyoga, defined at 11 (§9): Parasamaññā payoge ), and the rule assigns the designation garu to the short letter that precedes it. garu is carried into the next rule. There is no option marker.
Dvinnaṃ samūho du, tasmiṃ dumhi. Saṃyogabhūte akkhare pare yo pubbo rassakkharo, so garusañño hoti. Yathā – datvā, hitvā, bhutvā.
du is a group of two; dumhi means “before such a [pair]”. The short letter that stands before a letter that is a conjunct (saṃyoga) receives the designation garu (heavy). For example: datvā, hitvā, bhutvā.
lemma
inflected
meaning
🔽(√dā + tvā)
datvā
having given — short da heavy before tv
🔽(√hā + tvā)
hitvā
having abandoned — short hi heavy before tv
🔽(√bhuja + tvā)
bhutvā
having eaten — short bhu heavy before tv
‘‘Garū’’ti vattate.
garu continues [into the next rule].
The companion of 6 (§602): Dumhi garu , with garu carried over: by saying ca the rule extends the designation to a second class, the syllable whose vowel is long, whatever follows it. Buddhappiya’s five examples run through the five long vowels in the order of 5 (§5): Aññe dīghā — ā in nāvā, ī in nadī, ū in vadhū, e in dve, o in tayo — and he adds the definition Kaccāyana leaves to be inferred: every other syllable, a short vowel not before a conjunct, is lahu (light). ca here is an extension, not an option marker.
Dīgho ca saro garusañño hoti. Yathā – nāvā, nadī, vadhū, dve, tayo. Garukato añño ‘‘lahuko’’ti veditabbo.
And a long vowel receives the designation garu. For example: nāvā, nadī, vadhū, dve, tayo. [A syllable] other than a heavy one is to be understood as lahuka (light).
lemma
inflected
meaning
🚺①⨀(nāvā)
nāvā
a boat — ā heavy
🚺①⨀(nadī)
nadī
a river — ī heavy
🚺①⨀(vadhū)
vadhū
a young woman — ū heavy
⚧️①⨂(dvi)
dve
two — e heavy
🚹①⨂(ti)
tayo
three — o heavy; ta before it light
Back in Kaccāyana’s sequence: the vowels being defined, the remaining letters are named byañjana. Where Kaccāyana’s vutti lists them, Buddhappiya also counts them (thirty-three), gives their measure (half a mattā, on the scale of 4 (§4): Lahumattā tayo rassā ), glosses sesā (“other than those already stated”), derives the name — they make meaning manifest, byañjīyati — and points out that the a in ka, kha and the rest is only a carrier for pronunciation: the consonant proper is the vowelless k, which is what 12 (§10): Pubbamadhoṭhitamassaraṃ sarena viyojaye will detach. byañjanā continues into the next rule. There is no option marker.
Ṭhapetvā aṭṭha sare sesā aḍḍhamattā akkharā ka kārādayo niggahītantā tettiṃsa byañjanā nāma honti. Vuttehi aññe sesā. Byañjīyati etehi atthoti byañjanā. Taṃ yathā – ka kha ga gha ṅa, ca cha ja jha ña, ṭa ṭha ḍa ḍha ṇa, ta tha da dha na, pa pha ba bha ma, ya ra lava sa ha ḷa aṃ-iti byañjanā. Ka kārādīsvakāro uccāraṇattho.
Setting aside the eight vowels, the remaining letters of half a measure, from ka to the niggahīta, thirty-three [in number], are called byañjana (consonants). “Remaining” means other than those already stated. They are byañjana because meaning is made manifest (byañjīyati) by them. Namely: ka kha ga gha ṅa, ca cha ja jha ña, ṭa ṭha ḍa ḍha ṇa, ta tha da dha na, pa pha ba bha ma, ya ra la va sa ha ḷa, aṃ — these are the consonants. The a in ka and the rest is for the sake of pronunciation.
class
letters
count
the five groups (vagga)
ka kha ga gha ṅa, ca cha ja jha ña, ṭa ṭha ḍa ḍha ṇa, ta tha da dha na, pa pha ba bha ma
25
the ungrouped (avagga)
ya ra la va sa ha ḷa
7
the niggahīta
aṃ
1
‘‘Byañjanā’’ti vattate.
byañjanā continues.
The definition of the groups, with byañjanā carried from 8 (§6): Sesā byañjanā . The rule statement is Kaccāyana’s; Buddhappiya adds the names in use — the ka-group, the ca-group and so on, after the first letter of each — a gloss of vagga as “collection” (samūha), and two analyses of the sutta’s pañcapañcaso: “by a division of five each” or “having five each”. mantā he analyses as “those whose end is ma”, which fixes the twenty-five as the letters from ka to ma and leaves ya and the rest outside the groups. There is no option marker.
Tesaṃ kho byañjanānaṃ ka kārādayo ma kārantā pañcavīsati byañjanā pañcapañcavibhāgena vaggā nāma honti. Taṃ yathā – ka kha ga gha ṅa, ca cha ja jha ña, ṭa ṭha ḍa ḍha ṇa, ta tha da dha na, pa pha ba bha ma-iti vaggā.
Of those consonants, the twenty-five from ka to ma, divided five by five, are called vagga (groups). Namely: ka kha ga gha ṅa, ca cha ja jha ña, ṭa ṭha ḍa ḍha ṇa, ta tha da dha na, pa pha ba bha ma — these are the groups.
group
first
second
third
fourth
fifth
ka-vaggaka
kha
ga
gha
ṅa
ca-vaggaca
cha
ja
jha
ña
ṭa-vaggaṭa
ṭha
ḍa
ḍha
ṇa
ta-vaggata
tha
da
dha
na
pa-vaggapa
pha
ba
bha
ma
Te pana paṭhamakkharavasena ka vaggaca vaggādivohāraṃ gatā, vaggo ti samūho, tattha pañcapañcavibhāgenāti vā pañca pañca etesamatthīti vā pañcapañcaso, mo anto yesaṃ te mantā.
These have come to be called, after their first letter, the ka-group, the ca-group and so on. vagga means a collection. In [the sutta], pañcapañcaso means either “by a division of five each” or “those which have five each”; mantā are those whose end is ma.
The last letter to be named, in Kaccāyana’s place. Kaccāyana’s vutti simply equates aṃ with niggahīta; Buddhappiya explains how the sutta manages to name a sound that cannot be pronounced alone: the a is a carrier for pronunciation, as in ka at 8 (§6): Sesā byañjanā , and iti points to what has just been uttered, so that the name attaches to the dot (bindu) after the a. He then gives two etymologies from gayhati “is pronounced” — pronounced in dependence on a short vowel (it never follows a long one, a fact he will use at 57 (§37): Niggahitañca ), or pronounced with the organ of articulation “checked” (niggahetvā) — and fixes the second in a verse. The closing sentence introduces the next rule as a paribhāsā. There is no option marker.
A kāro uccāraṇattho, iti saddo panānantaravuttanidassanattho, aṃiti yaṃ a kārato paraṃ vuttaṃ bindu, taṃ niggahītaṃ nāma hoti. Rassassaraṃ nissāya gayhati, karaṇaṃ niggahetvā gayhatīti vā niggahītaṃ.
The a is for pronunciation; the word iti serves to point to what has just been uttered: the dot (bindu) that is uttered after the a in aṃ is called niggahīta. It is niggahīta because it is pronounced (gayhati) in dependence on a short vowel, or because it is pronounced with the organ [of articulation] checked (niggahetvā).
Karaṇaṃ niggahetvāna, mukhenāvivaṭena yaṃ;
That which is spoken with the organ [of articulation] checked and with the mouth not opened niggahīta: the dot pronounced in the wake of a vowel.
Idha avuttānaṃ parasamaññānampi payojane sati gahaṇatthaṃ paribhāsamāha.
Now, so that the technical terms of other [grammars] that are not stated here may also be taken up when there is a use for them, he states a paribhāsā (an interpretive rule).
The rule that closes the Saññāvidhāna, where Kaccāyana also has it: the terms the grammar has not defined itself may be borrowed from “others” when there is a use for them (payoge sati). Kaccāyana’s vutti names the Sanskrit treatise and admits two terms, ghosa and aghosa, which 42 (§29): Vagge ghosāghosānaṃ tatiyapaṭhamā will need. Buddhappiya widens the source to “other grammarians” and turns the rule into a glossary of everything his own commentary has used or will use: ghosa and aghosa with their counts; dhanita and sithila, already used at 1 (§1): Attho akkharasaññāto ; lopa, the operation of 13 (§12): Sarā sare lopaṃ ; savaṇṇa, equated with sarūpa, the words of 15 (§13): Vā paro asarūpā and 16 (§14): Kvacāsavaṇṇaṃ lutte ; saṃyoga, used at 6 (§602): Dumhi garu ; and liṅga and pada, the stem and the inflected word, which 19 (§17): Yamedantassādeso (padanta) and the verse at 29 (§58): Ivaṇṇuvaṇṇā jhalā presuppose. payoge is the only condition; there is no option marker.
Yā ca pana parasmiṃ sakkataganthe, paresaṃ vā veyyākaraṇānaṃ samaññā ghosāghosalopasavaṇṇasaṃyogaliṅgādikā, tā payoge sati etthāpi yujjante.
And whatever technical terms (samaññā) there are in another [treatise], the Sanskrit book, or belonging to other grammarians — ghosa, aghosa, lopa, savaṇṇa, saṃyoga, liṅga and the rest — these are applied here too when there is a use for them.
Parasmiṃ, paresaṃ vā samaññā parasamaññā, veyyākaraṇe, veyyākaraṇamadhītānaṃ vā samaññātyattho. Payujjanaṃ payogo, viniyogo.
parasamaññā are terms in another [work], or terms of others — that is, terms in a grammar, or of those who have studied grammar. payoga is applying, employment.
Tattha vaggānaṃ paṭhamadutiyā, sa kāro ca aghosā. Vaggānaṃ tatiyacatutthapañcamā, ya ra lava ha ḷā cāti ekavīsati ghosā nāma.
Among these, the first and second [letters] of the groups, and sa, are aghosa (unvoiced). The third, fourth and fifth of the groups, and ya ra la va ha ḷa — twenty-one — are called ghosa (voiced).
Ettha ca vaggānaṃ dutiyacatutthā dhanitā tipi vuccanti, itare sithilā ti. Vināso lopo. Rassassarā sakadīghehi aññamaññaṃ savaṇṇā nāma, ye sarūpā tipi vuccanti. Sarānantaritāni byañjanāni saṃyogo. Dhātuppaccayavibhattivajjitamatthavaṃ liṅgaṃ. Vibhatyantaṃ padaṃ. Iccevamādi.
And here the second and fourth of the groups are also called dhanita (aspirated), the others sithila (unaspirated). lopa (elision) is disappearance. The short vowels together with their own long ones are mutually savaṇṇa (homogeneous), and these are also called sarūpa (of the same form). Consonants not separated by a vowel are a saṃyoga (conjunct). A meaningful [element] other than a root, a suffix or an inflection is a liṅga (stem). What ends in an inflection is a pada (word). And so on.
Iti saññāvidhānaṃ.
Thus the assignment of technical terms.
Atha sarasandhi vuccate.
Now vowel sandhi (sarasandhi) is stated.
Loka aggapuggalo, paññā indriyaṃ, tīṇi imāni, no hi etaṃ, bhikkhunī ovādo, mātu upaṭṭhānaṃ, sametu āyasmā, abhibhū āyatanaṃ, dhanā me atthi, sabbe eva, tayo assu dhammā, asanto ettha na dissanti itīdha sarādisaññāyaṃ sabbasandhikaraṇaṭṭhāne byañjanaviyojanatthaṃ paribhāsamāha.
Here, in loka aggapuggalo, paññā indriyaṃ, tīṇi imāni, no hi etaṃ, bhikkhunī ovādo, mātu upaṭṭhānaṃ, sametu āyasmā, abhibhū āyatanaṃ, dhanā me atthi, sabbe eva, tayo assu dhammā and asanto ettha na dissanti, now that the designations sara and the rest are known, he states a paribhāsā for the purpose of separating the consonant [from its vowel] at every place where sandhi is to be made.
Buddhappiya opens the Sarasandhividhāna with Kaccāyana’s first procedural rule but changes the order of the next two: Kaccāyana has separate (§10), rejoin (§11), elide (§12); the Rūpasiddhi has separate (12), elide (13), rejoin (14), so that the two paribhāsā frame the first operation in the order in which the work is actually done, and every derivation in the book runs §10, the operating rule, §11. The twelve phrases quoted in the opening sentence are the material: they are separated here, elided at 13 (§12): Sarā sare lopaṃ and rejoined at 14 (§11): Naye paraṃ yutte . His gloss reads the instrumental sarena in two ways — as a nissakka, “[apart] from the vowel”, or “together with its vowel” — and settles what the sutta never names: the thing detached is a consonant, which follows from assaraṃ, since only a consonant can be made “vowelless”. Where Kaccāyana’s vutti says sandhiṃ kattukāmo (“wishing to make sandhi”), his anatikkamanto adds that it is the consonant next to the junction, and no other, that is taken. There is no option marker: this is an instruction (vidhyaṅga), not an operation.
Sarenā ti nissakke karaṇavacanaṃ, sahayoge vā, sandhitabbe sarasahitaṃ pubbabyañjanaṃ anatikkamanto adhoṭhitamassarañca katvā sarato viyojayeti sarato byañjanaṃ viyojetabbaṃ. Ettha ca assara ggahaṇasāmatthiyena ‘‘byañjana’’nti laddhaṃ.
sarena is an instrumental form ③ used in the sense of separation (nissakka, “from the vowel”), or else in the sense of accompaniment (“with its vowel”). In [a pair of words] to be joined, not passing over the preceding consonant that carries its vowel, one should make it stand below (adhoṭhita) and vowelless (assara) and separate it from the vowel — that is, the consonant is to be detached from the vowel. And here, by the force of the word assara (“vowelless”), “consonant” is obtained [as the thing detached, though the sutta does not name it].
The general rule (utsarga) of vowel sandhi, placed between the two paribhāsā: a vowel before a vowel drops. Kaccāyana’s vutti states it in a line and gives examples; Buddhappiya’s gloss is a lesson in how a sutta is read. sarā names the thing operated on (kāriyī), sare the condition (nimitta), lopaṃ the operation (kāriya). The plural sarā tells us that before any one following vowel all sorts of vowels drop. The locative sare is a nimittasattamī, a locative of condition, and because the following vowel is the condition, no sandhi takes place when a letter or a pause (vaṇṇakālabyavadhāna) intervenes between the two. Most important for the rules that follow: a locative condition operates on what precedes it, not on what follows — so this rule elides the preceding vowel (pubbalopa), and 15 (§13): Vā paro asarūpā , which needs the word paro to elide the following one (paralopa), is the exception to it. There is no option marker: outside the exceptions, the elision is obligatory. The closing sentence marks assaraṃ and adhoṭhitaṃ as carried from 12 (§10): Pubbamadhoṭhitamassaraṃ sarena viyojaye into the next paribhāsā, “for smooth utterance”.
Sarā kho sabbepi sare pare ṭhite lopaṃ papponti. Lopo ti adassanaṃ anuccāraṇaṃ. Ettha sarā ti kāriyīniddeso. Bahuvacanaṃ panettha ekekasmiṃ sare pare bahūnaṃ lopañāpanatthaṃ. Sare ti nimittaniddeso, nimittasattamī cāyaṃ, nimittopādānasāmatthiyato vaṇṇakālabyavadhāne sandhikāriyaṃ na hoti. Lopa nti kāriyaniddeso, idaṃ pana suttaṃ upari paralopavidhānato pubbalopavidhānanti daṭṭhabbaṃ, evaṃ sabbattha sattamīniddese pubbasseva vidhi, na parassa vidhānanti veditabbaṃ.
Vowels, all of them, when a vowel stands next, undergo elision (lopa). lopa is non-appearance, non-pronunciation. Here sarā designates the thing operated on (kāriyī). The plural is used here so as to make known that before each single following vowel many [different] vowels are elided. sare designates the condition (nimitta), and this is a locative ⑦ of condition (nimittasattamī); by the force of taking [the following vowel] as the condition, the sandhi operation does not take place when a letter or a pause (vaṇṇa-kāla-byavadhāna) intervenes. lopaṃ designates the operation (kāriya). This sutta, moreover, is to be seen as prescribing the elision of the preceding [vowel], because the elision of the following one is prescribed further on; and so it is to be understood that everywhere, when the designation is in the locative, the operation is on the preceding [element] alone, and no operation is prescribed for the following one.
‘‘Assaraṃ, adhoṭhita’’nti ca vattate, siliṭṭhakathane paribhāsamāha.
assaraṃ and adhoṭhitaṃ also continue; for smooth utterance (siliṭṭhakathana) he [now] states a paribhāsā.
The second paribhāsā, moved after the elision so that it closes the procedure: the consonant left vowelless at 12 (§10): Pubbamadhoṭhitamassaraṃ sarena viyojaye is carried to the letter that now follows it, assaraṃ and adhoṭhitaṃ being carried over from the two preceding rules. Buddhappiya explains the sutta’s yutte, “where it is fitting”: it excludes the niggahīta, which is never carried to a following vowel — Kaccāyana’s vutti gives the same counter-example, akkocchi maṃ avadhi maṃ, without saying why. The twelve phrases of the section’s opening are completed here, so that the examples of the first three rules form one set. The closing sentence introduces the sixteen phrases of the next rule and the device by which its wording is obtained: the sare of 13 (§12): Sarā sare lopaṃ still governs, but by the maxim “the inflection changes according to the sense” (atthavasā vibhattivipariṇāmo) the sarā … lopaṃ of that rule is re-read as saramhā ⑤ (“after a vowel”) … saro ① (“the vowel”) … lopaṃ, which is what Vā paro asarūpā needs. There is no option marker.
Sararahitaṃ kho byañjanaṃ adhoṭhitaṃ parakkharaṃ naye yutte ṭhāneti paranayanaṃ kātabbaṃ. Ettha yutta ggahaṇaṃ niggahītanisedhanatthaṃ, tena ‘‘akkocchi maṃ avadhi ma’’ntiādīsu paranayanasandeho na hoti.
A consonant deprived of its vowel, standing below, one should carry to the following letter where it is fitting (yutte): that is, the carrying-over to the following (paranayana) is to be done. Here the word yutta serves to exclude the niggahīta; hence in akkocchi maṃ avadhi maṃ and the like there is no doubt about the carrying-over [the ṃ is not carried to the vowel].
Lokaggapuggalo, paññindriyaṃ, tīṇimāni, nohetaṃ, bhikkhunovādo, mātupaṭṭhānaṃ, sametāyasmā, abhibhāyatanaṃ, dhanā matthi, sabbeva, tayassu dhammā, asantettha na dissanti.
The twelve phrases, joined:
words
elided by §12
carried over by §11
meaning
loka aggapuggalo
alokaggapuggalo the foremost person in the world
paññā indriyaṃ
āpaññindriyaṃ the faculty of wisdom
tīṇi imāni
itīṇimāni these three
no hi etaṃ
inohetaṃ certainly not
bhikkhunī ovādo
ībhikkhunovādo the exhortation of the nuns
mātu upaṭṭhānaṃ
umātupaṭṭhānaṃ attending on one’s mother
sametu āyasmā
usametāyasmā let the venerable one agree
abhibhū āyatanaṃ
ūabhibhāyatanaṃ the base of mastery
dhanā me atthi
edhanā matthi I have riches
sabbe eva
esabbeva all indeed
tayo assu dhammā
otayassu dhammā three things indeed
asanto ettha na dissanti
oasantettha na dissanti the bad are not seen here
tayassu dhammā jahitā bhavanti (18Kh 6: Ratanasutta)
asantettha na dissanti (18Dh 21.8: Cūḷasubhaddāvatthu)
sametāyasmā saṃghena (1V 1.2.10: Saṃghabhedasikkhāpada)
Yassa idāni, saññā iti, chāyā iva, kathā eva kā, iti api, assamaṇī asi, cakkhu indriyaṃ, akataññū asi, ākāse iva, te api, vande ahaṃ, so ahaṃ, cattāro ime, vasalo iti, moggallāno āsi bījako, pāto evātīdha pubbalope sampatte ‘‘sare’’ti adhikāro, idha pana ‘‘atthavasā vibhattivipariṇāmo’’ti katvā ‘‘saro, saramhā, lopa’’nti ca vattamāne –
Here, in yassa idāni, saññā iti, chāyā iva, kathā eva kā, iti api, assamaṇī asi, cakkhu indriyaṃ, akataññū asi, ākāse iva, te api, vande ahaṃ, so ahaṃ, cattāro ime, vasalo iti, moggallāno āsi bījako and pāto eva, when the elision of the preceding vowel [by the last rule] is due, the governing word sare [still holds]; but here, applying [the maxim] “the inflection changes according to the sense” (atthavasā vibhattivipariṇāmo), while saro [①, “the vowel”], saramhā [⑤, “after a vowel”] and lopaṃ continue –
The first exception to 13 (§12): Sarā sare lopaṃ , in Kaccāyana’s place: after a dissimilar vowel it is the following vowel that drops, and only optionally. Buddhappiya reads the sutta with the words of 13 re-inflected by the maxim quoted at the end of 14 (§11): Naye paraṃ yutte — saramhā ⑤ paro saro ① lopaṃ … vā, “after a vowel, the following vowel [undergoes] elision, or [not]”. Where Kaccāyana’s vutti gives two examples and two counter-examples, he gives sixteen examples, ten places where the rule does not reach, and a summary verse; and this is the sutta in which he sets out his doctrine of vā (Ruiz-Falqués, Boundaries and Domains §3.2–§3.3). Like the prefix ā, which marks either an exclusive limit (mariyādā, “up to but not including”) or an inclusive one (abhividhi, “up to and including”), vā has two uses: in some rules an open option (vikappa), in others the delimitation of forms as they are fixed in usage (yathāvavatthitarūpapariggaha) — the vavatthitavibhāsā, or restricted option. Here it is the latter, and the restricted option has three domains. nicca: the elision of the following vowel is the only sandhi possible — cakkhu indriyaṃ gives cakkhundriyaṃ and never cakkhindriyaṃ, Buddhappiya’s own niccaṃ; Ruiz-Falqués reads yassa idāni and the other idāni, iti, iva, eva cases the same way, since yassidāni does not occur. anicca: the elision is one of two correct forms — vasaloti or vasalo iti (the canon has the latter), moggallānosi or moggallāno āsi. asanta, “inapplicable”: the idha na bhavati list, where the following vowel never drops and the general rule 13 elides the preceding one instead (pañcindriyāni, never pañcandriyāni). The verse draws the boundary: after a or ā the following vowel is not elided except in idāni, iti, iva, eva and the like; after any other vowel a following long vowel is not elided, except in āsi, eva and the like. The last block carries paro and asarūpe into the next rule and, by a lion’s glance (sīhagati), fetches the words ivaṇṇa from 21 (§21): Ivaṇṇo yaṃ navā and u from 20 (§18): Vamodudantānaṃ to serve as its subject.
Asamānarūpamhā saramhā paro saro lopaṃ pappoti vā. Samānaṃ rūpaṃ assāti sarūpo, na sarūpo asarūpo, asavaṇṇo. Yasmā pana mariyādāyaṃ, abhividhimhi ca vattamāno ā upasaggo viya vā saddo dvidhā vattate, katthaci vikappe, katthaci yathāvavatthitarūpapariggahe, idha pana pacchime, tato niccamaniccamasantañca vidhimettha vā saddo dīpeti. ‘‘Naye paraṃ yutte’’ti paraṃ netabbaṃ.
After a vowel of dissimilar form, the following vowel optionally (vā) undergoes elision. sarūpa is that which has the same form; asarūpa is what is not sarūpa, that is, not homogeneous (asavaṇṇa). Now, just as the prefix ā, which is used for an exclusive limit (mariyādā) and for an inclusive limit (abhividhi), so the word vā is used in two ways: in some places for an open option (vikappa), in some places for the delimitation of forms as they are fixed [in usage] (yathāvavatthitarūpapariggaha); and here it is the latter. Hence in this rule the word vā indicates an operation that is [at once] mandatory (nicca), non-mandatory (anicca) and inapplicable (asanta). By 14 (§11): Naye paraṃ yutte the [consonant] is to be carried to the following.
Yassadāni yassa idāni, saññāti saññā iti, chāyāva chāyā iva, kathāva kā kathā eva kā, itīpi iti api, assamaṇīsi assamaṇī asi. Cakkhundriya miti niccaṃ. Akataññūsi akataññū asi, ākāseva ākāse iva, tepi te api, vandehaṃ vande ahaṃ, sohaṃ so ahaṃ, cattārome cattāro ime, vasaloti vasalo iti, moggallānosi bījako moggallāno āsi bījako, pātova pāto eva.
yassadāni from yassa idāni, saññāti from saññā iti, chāyāva from chāyā iva, kathāva kā from kathā eva kā, itīpi from iti api, assamaṇīsi from assamaṇī asi. cakkhundriyaṃ [from cakkhu indriyaṃ] — this one always (niccaṃ). akataññūsi from akataññū asi, ākāseva from ākāse iva, tepi from te api, vandehaṃ from vande ahaṃ, sohaṃ from so ahaṃ, cattārome from cattāro ime, vasaloti from vasalo iti, moggallānosi bījako from moggallāno āsi bījako, pātova from pāto eva.
Idha na bhavati – pañcindriyāni, saddhindriyaṃ, sattuttamo, ekūnavīsati, yassete, sugatovādo, diṭṭhāsavo, diṭṭhogho, cakkhāyatanaṃ, taṃ kutettha labbhā iccādi.
Here it does not apply: pañcindriyāni, saddhindriyaṃ, sattuttamo, ekūnavīsati, yassete, sugatovādo, diṭṭhāsavo, diṭṭhogho, cakkhāyatanaṃ, taṃ kutettha labbhā and so on.
By saying vā in the sense of a restricted option, the rule has an asanta domain, where the following vowel is never elided; in these words the preceding vowel drops by the general rule 13 (§12): Sarā sare lopaṃ , and where the survivor is i or u after a/ā it may then become e/o by 16 (§14): Kvacāsavaṇṇaṃ lutte or lengthen by 17 (§15): Dīghaṃ .
words
preceding vowel elided by §12
result
never
pañca indriyāni
apañcindriyāni pañcandriyāni
saddhā indriyaṃ
āsaddhindriyaṃ saddhāndriyaṃ
satta uttamo
asattuttamo sattattamo
eka ūna vīsati
aekūnavīsati ekanavīsati
yassa ete
ayassete yassate
sugata ovādo
asugatovādo sugatavādo
diṭṭhi āsavo
idiṭṭhāsavo diṭṭhisavo
diṭṭhi ogho
idiṭṭhogho diṭṭhigho
cakkhu āyatanaṃ
ucakkhāyatanaṃ cakkhuyatanaṃ
kuto ettha
okutettha kutottha
taṃ kutettha labbhā (1V Veranjakanda: Verañjakaṇḍa)
yassete (12S1 10.1.12: Āḷavakasutta)
Bhavati ca vavatthitavibhāsāya.
And there is [this summary verse] on the restricted option (vavatthitavibhāsā):
Avaṇṇato sarodānī-tīvevādiṃ vinā paro;
After an a-vaṇṇa the following vowel is not elided, except [the initial of] idāni, iti, iva, eva and the like; āsi, eva and the like.
Bandhussa iva, upa ikkhati, upa ito, ava icca, jina īritaṃ, na upeti, canda udayo, yathā udake itīdha pubbāvaṇṇassarānaṃ lope kate ‘‘paro, asarūpe’’ti ca vattate, tathā ‘‘ivaṇṇo yaṃ navā’’ti ito ivaṇṇa ggahaṇañca, ‘‘vamodudantāna’’nti ito u ggahaṇañca sīhagatiyā idhānuvattetabbaṃ.
Here, in bandhussa iva, upa ikkhati, upa ito, ava icca, jina īritaṃ, na upeti, canda udayo and yathā udake, when the elision of the preceding a-vaṇṇa vowel has been made, paro and asarūpe continue [into the next rule]; likewise the word ivaṇṇa from 21 (§21): Ivaṇṇo yaṃ navā and the word u from 20 (§18): Vamodudantānaṃ are to be carried here by a lion’s glance (sīhagati, a later rule looking back).
The second exception, in Kaccāyana’s place: once the preceding vowel has dropped by 13 (§12): Sarā sare lopaṃ , a surviving i, ī or u may become e or o. Buddhappiya’s gloss names the mechanism — e and o are the asavaṇṇa, the vowels that have no homogeneous partner, and each is the vowel nearest in place of articulation to the one it replaces (ṭhānāsannavasena) — and adds two things the Kaccāyana vutti does not say. First, he assembles the rule’s subject from two later rules by a lion’s glance, as announced at the end of 15 (§13): Vā paro asarūpā : ivaṇṇa from 21 (§21): Ivaṇṇo yaṃ navā and u (not u-vaṇṇa: ū never becomes o) from 20 (§18): Vamodudantānaṃ . Second, he limits the rule to the case where the elided vowel was a or ā: the vā of 15 is still in force, but the use of kvaci cancels it here (kvaci karaṇato), so diṭṭhi upādānaṃ gives diṭṭhupādānaṃ, not diṭṭhopādānaṃ, and yo me issaro gives yo missaro. kvaci is the exception marker of Boundaries and Domains §4.2: it isolates from the general state of affairs the words in which the change does happen (bandhusseva, upekkhati, candodayo) and leaves tatrime, yassindriyāni, lokuttaro to the plain elision of 13. The three ti kiṃ? questions test kvaci, lutte and asarūpe in turn. The last two blocks list the twenty-two phrases of the next rule and carry kvaci (as an adhikāra), paro and lutte into it.
I vaṇṇabhūto, u kārabhūto ca paro saro asarūpe pubbassare lutte kvaci asavaṇṇaṃ pappoti. Natthi savaṇṇā etesanti asavaṇṇā, ekārokārā, tattha ṭhānāsannavasena ivaṇṇukārānamekārokārā honti.
A following vowel that is an i-vaṇṇa, or that is u, when the preceding dissimilar vowel has been elided, in some places (kvaci) becomes the non-homogeneous [vowel]. asavaṇṇa are those that have no homogeneous [partner], namely e and o; there, by nearness of place [of articulation], i-vaṇṇa becomes e and u becomes o.
Bandhusseva, upekkhati, upeto, avecca, jineritaṃ, nopeti, candodayo, yathodake.
bandhusseva, upekkhati, upeto, avecca, jineritaṃ, nopeti, candodayo, yathodake.
words
elided by §12
changed by §14
result
meaning
bandhussa iva
ai→e
bandhusseva like a kinsman’s
upa ikkhati
ai→e
upekkhati he looks on
upa ito
ai→e
upeto endowed with
ava icca
ai→e
avecca having understood
jina īritaṃ
aī→e
jineritaṃ spoken by the Conqueror
na upeti
au→o
nopeti it does not approach
canda udayo
au→o
candodayo moonrise
yathā udake
āu→o
yathodake as in water
bandhussa + iva
rule
→
bandhuss + (a) + (i) + va
§10
→
bandhuss + (a + i) + va
§12
→
bandhuss + (i→e) + va
§14
→
(bandhuss + e) + va
§11
→
bandhusseva
canda + udayo
rule
→
cand + (a) + (u) + dayo
§10
→
cand + (a + u) + dayo
§12
→
cand + (u→o) + dayo
§14
→
(cand + o) + dayo
§11
→
candodayo
yathodake āvile appasanne (22J 2.4.5: Anabhiratijātaka)
Kvacī ti kiṃ? Tatrime, yassindriyāni, mahiddhiko, sabbītiyo, tenupasaṅkami, lokuttaro. Lutte ti kiṃ? Cha ime dhammā, yathā idaṃ, kusalassa upasampadā. Asarūpe ti kiṃ? Cattārimāni, mātupaṭṭhānaṃ.
Why kvaci? [Because of] tatrime, yassindriyāni, mahiddhiko, sabbītiyo, tenupasaṅkami, lokuttaro. Why lutte? [Because of] cha ime dhammā, yathā idaṃ, kusalassa upasampadā. Why asarūpe? [Because of] cattārimāni, mātupaṭṭhānaṃ.
By saying kvaci (in some places) the change is not made everywhere the preceding a/ā has dropped: here the surviving i or u stays as it is.
words
result
not
tatra ime
tatrime tatreme
yassa indriyāni
yassindriyāni yassendriyāni
mahā iddhiko
mahiddhiko maheddhiko
sabba ītiyo
sabbītiyo sabbetiyo
tena upasaṅkami
tenupasaṅkami tenopasaṅkami
loka uttaro
lokuttaro lokottaro
yassindriyāni samathaṅgatāni (18Dh 7.5: Mahākaccāyanattheravatthu)
By saying lutte (when [the preceding vowel] has been elided) there is no change where no elision has taken place: cha ime dhammā, yathā idaṃ, kusalassa upasampadā stand unjoined.
kusalassa upasampadā (7D 1.16: Cārikāanujānana)
By saying asarūpe (a dissimilar [vowel]) there is no change when the two vowels were homogeneous: in cattāri imāni and mātu upaṭṭhānaṃ the preceding i and u drop by 13 (§12): Sarā sare lopaṃ and the survivors, cattārimāni and mātupaṭṭhānaṃ, keep their form.
Ettha ca satipi heṭṭhā vā ggahaṇe kvaci karaṇato avaṇṇe eva lutte idha vuttavidhi hotīti daṭṭhabbaṃ. Tato idha na bhavati – diṭṭhupādānaṃ, pañcahupāli, mudindriyaṃ, yo missaroti.
And here, although the word vā stands in the rule before [and continues], because kvaci is used it is to be understood that the operation stated here takes place only when an a-vaṇṇa has been elided. Hence it does not apply here: diṭṭhupādānaṃ, pañcahupāli, mudindriyaṃ, yo missaro.
Tatra ayaṃ, buddha anussati, sa atthikā, paññavā assa, tadā ahaṃ, yāni idha bhūtāni, gacchāmi iti, ati ito, kikī iva, bahu upakāraṃ, madhu udakaṃ, su upadhāritaṃ, yopi ayaṃ, idāni ahaṃ, sace ayaṃ, appassuto ayaṃ, itara itarena, saddhā idha vittaṃ, kamma upanissayo, tathā upamaṃ, ratti uparato, vi upasamo iccatra pubbassarānaṃ lope kate –
Here, in tatra ayaṃ, buddha anussati, sa atthikā, paññavā assa, tadā ahaṃ, yāni idha bhūtāni, gacchāmi iti, ati ito, kikī iva, bahu upakāraṃ, madhu udakaṃ, su upadhāritaṃ, yopi ayaṃ, idāni ahaṃ, sace ayaṃ, appassuto ayaṃ, itara itarena, saddhā idha vittaṃ, kamma upanissayo, tathā upamaṃ, ratti uparato and vi upasamo, when the elision of the preceding vowels has been made –
‘‘Kvacī’’ti adhikāro, ‘‘paro, lutte’’ti ca vattate.
kvaci is the governing word (adhikāra); paro and lutte also continue.
The third exception, again in Kaccāyana’s place: after the preceding vowel has dropped, the following one may be lengthened. Buddhappiya gives the mechanism, parallel to 16 (§14): Kvacāsavaṇṇaṃ lutte : by nearness of place each short vowel takes its own homogeneous long (savaṇṇadīgha) — a to ā, i to ī, u to ū. kvaci, carried from 16 as the governing word, is again the exception marker: the twenty-two examples lengthen, the thirteen Kvacī ti kiṃ? cases do not (tīṇimāni, pañcaṅgiko, munindo, gacchāmahaṃ …), and nothing but usage decides which is which — that is what kvaci says (Boundaries and Domains §4.1–§4.2). The Lutte ti kiṃ? cases show that without an elision there is no lengthening: yathā ayaṃ, kikī iva stand unjoined. Kaccāyana’s vutti has three examples; Buddhappiya’s list runs through the elided vowels (a, ā, i, ī, u, e, o) and the vowels lengthened (a, i, u). The last two blocks list the nine phrases of the next rule, in which it is the following vowel that has dropped, and carry lutte and dīghaṃ into it.
Saro kho paro pubbassare lutte kvaci dīghabhāvaṃ pappotīti ṭhānāsannavasena rassassarānaṃ savaṇṇadīgho.
A following vowel, when the preceding vowel has been elided, in some places becomes long — that is, by nearness of place [of articulation] the short vowels take their homogeneous long (savaṇṇadīgha).
Tatrāyaṃ, buddhānussati, sātthikā, paññavāssa, tadāhaṃ, yānīdha bhūtāni, gacchāmīti, atīto, kikīva, ba hūpakāraṃ, madhūdakaṃ, sūpadhāritaṃ, yopāyaṃ, idānāhaṃ, sacāyaṃ, appassutāyaṃ, itarītarena, saddhīdha vittaṃ, kammūpanissayo, tathūpamaṃ, rattūparato, vūpasamo.
tatrāyaṃ, buddhānussati, sātthikā, paññavāssa, tadāhaṃ, yānīdha bhūtāni, gacchāmīti, atīto, kikīva, bahūpakāraṃ, madhūdakaṃ, sūpadhāritaṃ, yopāyaṃ, idānāhaṃ, sacāyaṃ, appassutāyaṃ, itarītarena, saddhīdha vittaṃ, kammūpanissayo, tathūpamaṃ, rattūparato, vūpasamo.
words
elided by §12
lengthened by §15
result
meaning
tatra ayaṃ
aa→ā
tatrāyaṃ there this
buddha anussati
aa→ā
buddhānussati recollection of the Buddha
sa atthikā
aa→ā
sātthikā with a purpose
paññavā assa
āa→ā
paññavāssa let him be wise
tadā ahaṃ
āa→ā
tadāhaṃ then I
yāni idha bhūtāni
ii→ī
yānīdha bhūtāni whatever beings here
gacchāmi iti
ii→ī
gacchāmīti “I go”
ati ito
ii→ī
atīto past
kikī iva
īi→ī
kikīva like a blue jay
bahu upakāraṃ
uu→ū
bahūpakāraṃ of great help
madhu udakaṃ
uu→ū
madhūdakaṃ honey-water
su upadhāritaṃ
uu→ū
sūpadhāritaṃ well considered
yopi ayaṃ
ia→ā
yopāyaṃ and whoever this
idāni ahaṃ
ia→ā
idānāhaṃ now I
sace ayaṃ
ea→ā
sacāyaṃ if this
appassuto ayaṃ
oa→ā
appassutāyaṃ this one of little learning
itara itarena
ai→ī
itarītarena with one or the other
saddhā idha vittaṃ
āi→ī
saddhīdha vittaṃ faith is wealth here
kamma upanissayo
au→ū
kammūpanissayo the support of kamma
tathā upamaṃ
āu→ū
tathūpamaṃ like that
ratti uparato
iu→ū
rattūparato abstaining at night
vi upasamo
iu→ū
vūpasamo calming
tatra + ayaṃ
rule
→
tatr + (a) + (a) + yaṃ
§10
→
tatr + (a + a) + yaṃ
§12
→
tatr + (a→ā) + yaṃ
§15
→
(tatr + ā) + yaṃ
§11
→
tatrāyaṃ
saddhā + idha
rule
→
saddh + (ā) + (i) + dha
§10
→
saddh + (ā + i) + dha
§12
→
saddh + (i→ī) + dha
§15
→
(saddh + ī) + dha
§11
→
saddhīdha
yānīdha bhūtāni samāgatāni (18Kh 6: Ratanasutta)
saddhīdha vittaṃ purisassa seṭṭhaṃ (12S1 1.8.3: Vittasutta)
kikīva aṇḍaṃ rakkheyya (20Ap1 2.6: Rāhulattheraapadāna)
paññavāssa yathā mūgo (19Th1 8.1.1: Mahākaccāyanattheragāthā)
Kvacī ti kiṃ? Aciraṃ vata’yaṃ kāyo, kimpimāya, tīṇimāni, pañcasupādānakkhandhesu, tassattho, pañcaṅgiko, munindo, satindriyaṃ, lahuṭṭhānaṃ, gacchāmahaṃ, tatridaṃ, pañcahupāli, natthaññaṃ. Lutte ti kiṃ? Yathā ayaṃ, nimi iva rājā, kikī iva, su upadhāritaṃ.
Why kvaci? [Because of] aciraṃ vatayaṃ kāyo, kimpimāya, tīṇimāni, pañcasupādānakkhandhesu, tassattho, pañcaṅgiko, munindo, satindriyaṃ, lahuṭṭhānaṃ, gacchāmahaṃ, tatridaṃ, pañcahupāli, natthaññaṃ. Why lutte? [Because of] yathā ayaṃ, nimi iva rājā, kikī iva, su upadhāritaṃ.
By saying kvaci (in some places) the lengthening is not general: in these words the preceding vowel has dropped and the survivor keeps its short measure.
words
result
not
aciraṃ vata ayaṃ kāyo
aciraṃ vatayaṃ kāyo vatāyaṃ
kiṃ pi imāya
kimpimāya kimpīmāya
tīṇi imāni
tīṇimāni tīṇīmāni
pañcasu upādānakkhandhesu
pañcasupādānakkhandhesu pañcasūpādāna°
tassa attho
tassattho tassāttho
pañca aṅgiko
pañcaṅgiko pañcāṅgiko
muni indo
munindo munīndo
sati indriyaṃ
satindriyaṃ satīndriyaṃ
lahu uṭṭhānaṃ
lahuṭṭhānaṃ lahūṭṭhānaṃ
gacchāmi ahaṃ
gacchāmahaṃ gacchāmāhaṃ
tatra idaṃ
tatridaṃ tatrīdaṃ
pañcahi upāli
pañcahupāli pañcahūpāli
natthi aññaṃ
natthaññaṃ natthāññaṃ
aciraṃ vatayaṃ kāyo (18Dh 3.7: Pūtigattatissattheravatthu)
kimpimāya aññaṃ kiñci adeyyaṃ bhavissati (3V 6.9: Manussamaṃsapaṭikkhepakathā)
By saying lutte (when [the preceding vowel] has been elided) there is no lengthening where nothing has been elided: yathā ayaṃ, nimi iva rājā, kikī iva and su upadhāritaṃ stand as they are.
Lokassa iti, deva iti, vi ati patanti, vi ati nāmenti, saṅghāṭi api, jīvitahetu api, vijju iva, kiṃsu idha vittaṃ, sādhu iti itīdha parassarānaṃ lope kate –
Here, in lokassa iti, deva iti, vi ati patanti, vi ati nāmenti, saṅghāṭi api, jīvitahetu api, vijju iva, kiṃsu idha vittaṃ and sādhu iti, when the elision of the following vowels has been made –
‘‘Lutte, dīgha’’nti ca vattate.
lutte and dīghaṃ also continue.
The mirror image of 17 (§15): Dīghaṃ , in Kaccāyana’s place: when the following vowel has dropped — by 15 (§13): Vā paro asarūpā , which is why every second word in the nine phrases begins with i, a or e of iti, api, idha, iva or ati — the preceding vowel may lengthen. Buddhappiya reads the sutta’s ca not as an extension but as anukaḍḍhana, “dragging in”: it pulls lutte and dīghaṃ down from the rule before, and its purpose is to signal the maxim that what ca drags in goes no further (cānukaḍḍhitaṃ uttaratra nānuvattate), so that lutte and dīghaṃ stop here and are not available to 19 (§17): Yamedantassādeso , whose lengthening he will take from the consonant-sandhi rule 37 (§25): Dīghaṃ instead. kvaci, still the governing word, is again the exception marker: lokassāti, vītipatanti, kiṃsūdha vittaṃ lengthen, but yassadāni, itissa, idānipi, tesupi, cakkhundriyaṃ, kinnumāva do not, though the following vowel has dropped in each. The closing block lists the five phrases of the next rule, in which the elision of the preceding vowel would be due.
Pubbo saro parassare lutte kvacidīghaṃ pappoti. Ca ggahaṇaṃ luttadīghaggahaṇānukaḍḍhanatthaṃ, taṃ ‘‘cā nukaḍḍhitamuttaratra nānuvattate’’ti ñāpanatthaṃ.
The preceding vowel, when the following vowel has been elided, in some places becomes long. The word ca is for dragging in (anukaḍḍhana) the words lutte and dīghaṃ [from the rules before], and that in order to make known [the maxim] “what is dragged in by ca does not continue further on”.
Lokassāti, devāti, vītipatanti, vītināmenti, saṅghāṭīpi, jīvitahetūpi, vijjūva, kiṃsūdha vittaṃ, sādhūti.
lokassāti, devāti, vītipatanti, vītināmenti, saṅghāṭīpi, jīvitahetūpi, vijjūva, kiṃsūdha vittaṃ, sādhūti.
words
following vowel elided by §13
lengthened by §16
result
meaning
lokassa iti
ia→ā
lokassāti “of the world”
deva iti
ia→ā
devāti “O king”
vi ati patanti
ai→ī
vītipatanti they flit about
vi ati nāmenti
ai→ī
vītināmenti they pass [the time]
saṅghāṭi api
ai→ī
saṅghāṭīpi the outer robe too
jīvitahetu api
au→ū
jīvitahetūpi even for life’s sake
vijju iva
iu→ū
vijjūva like lightning
kiṃsu idha vittaṃ
iu→ū
kiṃsūdha vittaṃ what here is wealth
sādhu iti
iu→ū
sādhūti “good”
lokassa + iti
rule
→
lokass + (a) + (i) + ti
§10
→
lokass + (a + i) + ti
§13
→
lokass + (a→ā) + ti
§16
→
(lokass + ā) + ti
§11
→
lokassāti
kiṃsu + idha
rule
→
kiṃs + (u) + (i) + dha
§10
→
kiṃs + (u + i) + dha
§13
→
kiṃs + (u→ū) + dha
§16
→
(kiṃs + ū) + dha
§11
→
kiṃsūdha
vītipatanti (18Sn 3.11: Nālakasutta)
kiṃsūdha vittaṃ purisassa seṭṭhaṃ (12S1 1.8.3: Vittasutta)
jīvitahetūpi (21Cp 2.1.5: Mahiṃsarājacariya)
vijjūva (21Bu 17: Dhammadassībuddhavaṃsa)
Kvacī ti kiṃ? Yassadāni, itissa, idānipi, tesupi, cakkhundriyaṃ, kinnumāva.
Why kvaci? [Because of] yassadāni, itissa, idānipi, tesupi, cakkhundriyaṃ, kinnumāva.
Adhigato kho me ayaṃ dhammo, putto te ahaṃ, te assa pahīnā, pabbate ahaṃ, ye assa itīdha pubbalope sampatte –
Here, in adhigato kho me ayaṃ dhammo, putto te ahaṃ, te assa pahīnā, pabbate ahaṃ and ye assa, when the elision of the preceding vowel is due –
Kaccāyana’s §17 in its place, after the two lengthening rules. The sutta has no marker of its own; Buddhappiya writes kvaci into his gloss, the governing word carried from 16 (§14): Kvacāsavaṇṇaṃ lutte , because te anāgatā gives tenāgatā and puttā me atthi gives puttā matthi with no y — the exception marker again. What his gloss adds is a list: the operation applies only to me, te, ye (and, from the examples, pabbate) before an a — a restriction of the vavatthita kind stated by enumeration rather than by name. He analyses the sutta’s three words (yaṃ is the form ya; edanta “ending in e”; ādesa “what is directed to the place of another”) and explains the purpose of anta: an e inside a word, as in pavattento from pavatte + anta, is not replaced. The lengthening of tyāhaṃ he takes not from 17 (§15): Dīghaṃ — whose dīghaṃ the ca of 18 (§16): Pubbo ca has stopped — but from the consonant-sandhi rule 37 (§25): Dīghaṃ under the governing word byañjane: once te has become ty, the a of ahaṃ stands before the consonant h and lengthens there. (The Kaccāyana page derives the same forms by §15.) The closing block lists the eighteen phrases of the next rule.
Ekārassa padantabhūtassa ṭhāne sare pare kvaci ya kārādeso hoti. A kāremeteyesaddādissevāyaṃ vidhi, ya nti yaṃ rūpaṃ, e eva anto edanto, ādesiṭṭhāne ādissatīti ādeso. ‘‘Byañjane’’ti adhikicca ‘‘dīgha’’nti dīgho.
In place of an e that stands at the end of a word, when a vowel follows, there is in some places (kvaci) the replacement ya. This operation applies only to words such as me, te and ye before an a. yaṃ means the form ya; edanta is that whose end is e itself; ādesa (replacement) is what is directed into the place of the thing replaced. The lengthening is by 37 (§25): Dīghaṃ under the governing word byañjane.
Adhigato kho myāyaṃ dhammo, putto tyāhaṃ, tyāssa pahīnā, pabbatyāhaṃ, yyāssa.
adhigato kho myāyaṃ dhammo, putto tyāhaṃ, tyāssa pahīnā, pabbatyāhaṃ, yyāssa.
words
e→y by §17lengthened by §25
result
meaning
me ayaṃ
my + ayaṃ
a→ā
myāyaṃ by me this
te ahaṃ
ty + ahaṃ
a→ā
tyāhaṃ your … I
te assa
ty + assa
a→ā
tyāssa they … of him
pabbate ahaṃ
pabbaty + ahaṃ
a→ā
pabbatyāhaṃ on the mountain I
ye assa
yy + assa
a→ā
yyāssa those who … of him
me + ayaṃ
rule
→
(m + e) + (a) + yaṃ
§10
→
m + (e→y) + a + yaṃ
§17
→
m + y + (a→ā) + yaṃ
§25
→
(m + y + ā) + yaṃ
§11
→
myāyaṃ
te + ahaṃ
rule
→
(t + e) + (a) + haṃ
§10
→
t + (e→y) + a + haṃ
§17
→
t + y + (a→ā) + haṃ
§25
→
(t + y + ā) + haṃ
§11
→
tyāhaṃ
adhigato kho myāyaṃ dhammo (3V 1.5: Brahmayācanakathā)
putto tyāhaṃ mahārāja (22J 1.1.7: Kaṭṭhahārijātaka)
tyāssa pahīnā honti (11M 4.10: Dhātuvibhaṅgasutta)
pabbatyāhaṃ (23J 22.1.3: Suvaṇṇasāmajātaka)
yyāssa diṭṭhānugatiṃ āpajjeyyuṃ (15A3 2.4.11: Saṅkavāsutta)
Kvacī ti kiṃ? Te nāgatā, puttā matthi. Anta ggahaṇaṃ kiṃ? Dhammacakkaṃ pavattento, damento cittaṃ.
Why kvaci? [Because of] te nāgatā, puttā matthi. Why the word anta? [Because of] dhammacakkaṃ pavattento, damento cittaṃ.
By saying kvaci (in some places) a final e before a vowel is not always replaced: in te anāgatā the following a drops by 15 (§13): Vā paro asarūpā and gives tenāgatā; in puttā me atthi the e itself drops by 13 (§12): Sarā sare lopaṃ and gives puttā matthi.
puttā matthi dhanaṃ matthi (18Dh 5.3: Ānandaseṭṭhivatthu)
By saying anta (at the end [of a word]) an e inside a word is not replaced: in pavattento (pavatte + anta) and damento (dame + anta) the e before the a of the suffix stays, and the a drops.
dhammacakkaṃ pavattento (20Ap1 40.1.6: Dānakathā)
Yāvatako assa kāyo, tāvatako assa byāmo, ko attho, atha kho assa, ahaṃ kho ajja, yo ayaṃ, so assa, so eva, yato adhikaraṇaṃ, anu addhamāsaṃ, anu eti, su āgataṃ, su ākāro, du ākāro, cakkhu āpāthaṃ, bahu ābādho, pātu akāsi, na tu evātīdha –
Here, in yāvatako assa kāyo, tāvatako assa byāmo, ko attho, atha kho assa, ahaṃ kho ajja, yo ayaṃ, so assa, so eva, yato adhikaraṇaṃ, anu addhamāsaṃ, anu eti, su āgataṃ, su ākāro, du ākāro, cakkhu āpāthaṃ, bahu ābādho, pātu akāsi and na tu eva –
Kaccāyana’s §18 in its place, the counterpart of 19 (§17): Yamedantassādeso for a final o or u. kvaci still governs and marks the exception: ko attho and atha kho aññatarā stay unjoined, yohaṃ, sohaṃ, cattārome take the elision of 15 (§13): Vā paro asarūpā , sāgataṃ and sādhāvuso that of 13 (§12): Sarā sare lopaṃ , and v appears only where usage has it. Buddhappiya’s gloss again names the words: the o meant is the o of ko, kho, yo and so — the o of the pronouns and particles, not every final o — while the u of the examples belongs to the prefixes and particles anu, su, du, pātu, tu and to the stems cakkhu and bahu. He explains anta (a u inside a word, as in savanīyaṃ and viravanti, is not replaced) and in the closing block carries asarūpe from 15 to the next rule by a frog’s leap (maṇḍūkagati) over the four rules between. One reading of the CST is corrected.
Okārukārānaṃ antabhūtānaṃ sare pare kvaci va kārādeso hoti. Ka kha ya ta saddādio kārassedaṃ gahaṇaṃ.
For an o or u that stands at the end [of a word], when a vowel follows, there is in some places (kvaci) the replacement va. This [o] means the o of words such as ko, kho, yo and so [the pronouns ka, kha, ya, ta].
Yāvatakvassa kāyo, tāvatakvassa byāmo, kvattho, atha khvassa, ahaṃ khvajja, yvāyaṃ, svassa, sveva, yatvādhikaraṇaṃ, anvaddhamāsaṃ, anvebhi2
yāvatakvassa kāyo, tāvatakvassa byāmo, kvattho, atha khvassa, ahaṃ khvajja, yvāyaṃ, svassa, sveva, yatvādhikaraṇaṃ, anvaddhamāsaṃ, anveti, svāgataṃ, svākāro, dvākāro, cakkhvāpāthaṃ, bahvābādho, pātvākāsi, na tveva.
words
replaced by §18
result
meaning
yāvatako assa kāyo
o→v
yāvatakvassa kāyo as long as his body
tāvatako assa byāmo
o→v
tāvatakvassa byāmo so long his arm-span
ko attho
o→v
kvattho what is the point
atha kho assa
o→v
atha khvassa and then his
ahaṃ kho ajja
o→v
ahaṃ khvajja I indeed today
yo ayaṃ
o→v
yvāyaṃ whoever this
so assa
o→v
svassa that of his
so eva
o→v
sveva that very
yato adhikaraṇaṃ
o→v
yatvādhikaraṇaṃ on account of which
anu addhamāsaṃ
u→v
anvaddhamāsaṃ every half-month
anu eti
u→v
anveti he follows
su āgataṃ
u→v
svāgataṃ welcome
su ākāro
u→v
svākāro of good disposition
du ākāro
u→v
dvākāro of bad disposition
cakkhu āpāthaṃ
u→v
cakkhvāpāthaṃ the range of the eye
bahu ābādho
u→v
bahvābādho sickly
pātu akāsi
u→v
pātvākāsi he made manifest
na tu eva
u→v
na tveva but not
kho + assa
rule
→
(kh + o) + (a) + ssa
§10
→
kh + (o→v) + a + ssa
§18
→
(kh + v + a) + ssa
§11
→
khvassa
pātu + akāsi
rule
→
pāt + (u) + (a) + kāsi
§10
→
pāt + (u→v) + a + kāsi
§18
→
pāt + v + (a→ā) + kāsi
§25
→
(pāt + v + ā) + kāsi
§11
→
pātvākāsi
In svāgataṃ, svākāro, dvākāro, cakkhvāpāthaṃ and bahvābādho the second word already begins with ā; in pātvākāsi and yatvādhikaraṇaṃ the a after the new v is lengthened as at 19.
yāvatakvassa kāyo (7D 1.3: Dvattiṃsamahāpurisalakkhaṇā)
atha khvassa (13S3 1.2.4.7: Khemakasutta)
ahaṃ khvajja (15A3 2.2.10: Uposathasutta)
yatvādhikaraṇamenaṃ (6D 10.2: Samādhikkhandha)
anvaddhamāsaṃ (2V 1.5.3.1: Ovādasikkhāpada)
bahvābādho (4V 4.8: Adhikaraṇa)
pātvākāsi (1V 1.1.4.2: Vinītavatthu)
Kvacī ti kiṃ? Ko attho, atha kho aññatarā, yohaṃ, sohaṃ, cattārome, sāgataṃ, sādhāvuso, hotūti. Anta ggahaṇaṃ kiṃ? Savanīyaṃ, viravanti.
Why kvaci? [Because of] ko attho, atha kho aññatarā, yohaṃ, sohaṃ, cattārome, sāgataṃ, sādhāvuso, hotūti. Why the word anta? [Because of] savanīyaṃ, viravanti.
By saying kvaci (in some places) a final o or u before a vowel is not always replaced: the same words are also left unjoined or take one of the elisions.
words
result
how
ko attho
ko attho unjoined
atha kho aññatarā
atha kho aññatarā unjoined
yo ahaṃ
yohaṃ following a elided by §13
so ahaṃ
sohaṃ following a elided by §13
cattāro ime
cattārome following i elided by §13
su āgataṃ
sāgataṃ u elided by §12
sādhu āvuso
sādhāvuso u elided by §12
hotu iti
hotūti following i elided by §13, u lengthened by §16
sohaṃ (1V 1.1.1.1: Sudinnabhāṇavāra)
sādhāvuso (3V 4.25: Bhaṇḍanakārakavatthu)
idaṃ ayyassa hotūti (1V 1.4.2.8: Rūpiyasikkhāpada)
By saying anta (at the end [of a word]) a u inside a word is not replaced: in savanīyaṃ (√su + anīya) and viravanti (vi + √ru + a + nti) the u of the root does not become v by this rule.
Paṭisanthāravutti assa, sabbā vitti anubhuyyate, vi añjanaṃ, vi ākato, nadī āsanno itīdha maṇḍūkagatiyā ‘‘asarūpe’’ti vattate.
Here, in paṭisanthāravutti assa, sabbā vitti anubhuyyate, vi añjanaṃ, vi ākato and nadī āsanno, by a frog’s leap (maṇḍūkagati, over the rules between) asarūpe continues [from 15].
Kaccāyana’s §21, which Buddhappiya moves forward to follow §18 — §19 and §20 he places at 22 (§19): Sabbo canti and 27 (§20): Do dhassa ca — so that the three replacements of a final vowel stand together: e to y (19), o and u to v (20), i and ī to y (21). This is the sutta Ruiz-Falqués cites for Buddhappiya’s reading of navā (Boundaries and Domains §2.3 and §4.5): navā saddo kvaci saddapariyāyo, “the word navā is a synonym of the word kvaci” — an exception marker, not an open option, in agreement with Vimalabuddhi, for whom the navā of Kacc §21 blocks the kvaci carried from the rules before it, an exception to an exception. Buddhappiya adopts the equation without arguing for it; the Kaccāyana page of §21 reads the word more restrictively (“rarely”) and keeps it apart from kvaci. So paṭisanthāravutyassa, byañjanaṃ, nadyāsanno take y, while pañcahaṅgehi, gacchāmahaṃ take the elision of 13 (§12): Sarā sare lopaṃ and tāni attani, muttacāgī anuddhato stay unjoined. The second condition is one Kaccāyana’s vutti (sare pare) does not have: asarūpe, fetched from 15 (§13): Vā paro asarūpā by the frog’s leap announced at the end of 20 (§18): Vamodudantānaṃ — before a homogeneous i there is no y (itihidaṃ, aggīva, atthīti), the point on which the verse at 23 (§47): Atissa cantassa turns. The closing block lists the eight phrases of the next rule, in which y would be due and c is made instead.
Pubbo ivaṇṇo asarūpe sare pare ya kāraṃ pappoti navā. I eva vaṇṇo ivaṇṇo, navā saddo kvaci saddapariyāyo.
A preceding i-vaṇṇa, when a dissimilar vowel follows, becomes ya — navā. ivaṇṇa is the vaṇṇa that is i itself [i and ī]; the word navā is a synonym (pariyāya) of the word kvaci.
Paṭisanthāravutyassa, sabbā vityānubhuyyate, byañjanaṃ, byākato, nadyāsanno.
paṭisanthāravutyassa, sabbā vityānubhuyyate, byañjanaṃ, byākato, nadyāsanno.
Navā ti kiṃ? Pañcahaṅgehi, tāni attani, gacchāmahaṃ, muttacāgī anuddhato. Asarūpe ti kiṃ? Itihidaṃ, aggīva, atthīti.
Why navā? [Because of] pañcahaṅgehi, tāni attani, gacchāmahaṃ, muttacāgī anuddhato. Why asarūpe? [Because of] itihidaṃ, aggīva, atthīti.
By saying navā (= kvaci, in some places) a final i before a dissimilar vowel is not always replaced: in pañcahi aṅgehi and gacchāmi ahaṃ the i drops by 13 (§12): Sarā sare lopaṃ , and tāni attani, muttacāgī anuddhato stay unjoined.
words
result
how
pañcahi aṅgehi
pañcahaṅgehi i elided by §12
tāni attani
tāni attani unjoined
gacchāmi ahaṃ
gacchāmahaṃ i elided by §12
muttacāgī anuddhato
muttacāgī anuddhato unjoined
pañcahaṅgehi samannāgato (1V 1.4.2.8: Rūpiyasikkhāpada)
tāni attani passati (22J 2.8.2: Cūḷanandiyajātaka)
gacchāmahaṃ (3V 6.21: Meṇḍakagahapativatthu)
By saying asarūpe (before a dissimilar [vowel]) there is no y when the following vowel is i or ī: in iti hi idaṃ, aggi iva and atthi iti the first i drops by 13 (§12): Sarā sare lopaṃ — with lengthening by 17 (§15): Dīghaṃ in the last two — and the results are itihidaṃ, aggīva, atthīti.
aggīva (8D 8.12: Suhadamitta)
Ati antaṃ, ati odātā, pati ayo, pati āharati, pati eti, iti assa, iti etaṃ, itiādi itīdha ‘‘ivaṇṇo yaṃ navā’’ti ya kārādese sampatte –
Here, in ati antaṃ, ati odātā, pati ayo, pati āharati, pati eti, iti assa, iti etaṃ and iti ādi, when the replacement by ya by [this rule] Ivaṇṇo yaṃ navā is due –
Kaccāyana’s §19, which Buddhappiya takes out of its place between §18 and §20 and sets after §21, and he says why. The rule turns the whole syllable ti into c, so it must be stated for a ti whose i has not already become y by 21 (§21): Ivaṇṇo yaṃ navā — akatayakārassa. Read before 21, iti assa would first give ityassa and there would be no ti left for c to replace; and both the kvaci of this rule and the sutta 23 (§47): Atissa cantassa , which forbids the change before i, would be idle. The gloss also names the words meant (ati, pati, iti; the sutta’s sabbo says that the whole syllable, not the i alone, is replaced) and completes the derivation with the doubling rule of the consonant-sandhi section, 40 (§28): Paradvebhāvo ṭhāne : ati antaṃ gives ac + antaṃ and then accantaṃ. kvaci is the exception marker: iccassa, paccayo change, itissa, iti ākaṅkhamānena do not. The last block fetches na ivaṇṇe from 26 (§46): Te na vā ivaṇṇe , a rule not yet stated, for the sutta that follows.
Atipatiitīnaṃ ti saddassedaṃ gahaṇaṃ.
This [ti] means the sound ti of ati, pati and iti.
Sabbo ti icceso saddo sare pare kvaci ca kāraṃ pappoti. Tī ti niddesato akataya kārassevāyaṃ vidhi, itarathā kvaci ggahaṇassa ca ‘‘atissa cantassā’’tisuttassa ca niratthakatā siyā. ‘‘Paradvebhāvo ṭhāne’’ti dvittaṃ.
The whole of this sound ti, when a vowel follows, in some places (kvaci) becomes ca. Since the designation is ti, this operation is only for a ti whose [i] has not been made ya; otherwise the word kvaci and the sutta 23 (§47): Atissa cantassa would both be pointless. The doubling is by 40 (§28): Paradvebhāvo ṭhāne .
Accantaṃ, accodātā, paccayo, paccāharati, pacceti, iccassa, iccetaṃ, iccādi.
accantaṃ, accodātā, paccayo, paccāharati, pacceti, iccassa, iccetaṃ, iccādi.
words
ti→c by §19doubled by §28
result
meaning
ati antaṃ
ac + antaṃ
cc
accantaṃ utterly; the very end
ati odātā
ac + odātā
cc
accodātā too fair
pati ayo
pac + ayo
cc
paccayo a condition
pati āharati
pac + āharati
cc
paccāharati he brings back
pati eti
pac + eti
cc
pacceti he goes back to
iti assa
ic + assa
cc
iccassa thus to him
iti etaṃ
ic + etaṃ
cc
iccetaṃ thus this
iti ādi
ic + ādi
cc
iccādi and so on
ati + antaṃ
rule
→
a + (ti) + (a) + ntaṃ
§10
→
a + (ti→c) + a + ntaṃ
§19
→
(a + c→cc) + a + ntaṃ
§28
→
a + (cc + a) + ntaṃ
§11
→
accantaṃ
iti + etaṃ
rule
→
i + (ti) + (e) + taṃ
§10
→
i + (ti→c) + e + taṃ
§19
→
(i + c→cc) + e + taṃ
§28
→
i + (cc + e) + taṃ
§11
→
iccetaṃ
accodātā (2V 1.5.1.2: Omasavādasikkhāpada)
paccāharati (1V 1.2.5: Sañcarittasikkhāpada)
iccassa vacanīyaṃ (7D 2.1: Paṭiccasamuppāda)
iccetaṃ kusalaṃ (1V 1.2.10: Saṃghabhedasikkhāpada)
Kvacī ti kiṃ? Itissa, iti ākaṅkhamānena.
Why kvaci? [Because of] itissa, iti ākaṅkhamānena.
‘‘Te na vāivaṇṇe’’ti ito ‘‘na ivaṇṇe’’ti ca vattate.
From 26 (§46): Te na vā ivaṇṇe the words na and ivaṇṇe also continue [into the next rule].
Kaccāyana states this prohibition at §47, among the prefix rules at the end of the chapter; Buddhappiya brings it forward to stand beside the rule it limits, 22 (§19): Sabbo canti , with na and ivaṇṇe fetched from 26 (§46): Te na vā ivaṇṇe for the purpose. His gloss defends the sutta on two fronts. atissa is an anukaraṇa, the prefix quoted as a word, so its inflection stands (a genitive ⑥ that is not dropped). And anta must mean the ti at the end of ati, not its final i: a prohibition of y for that i would be empty, because 21 (§21): Ivaṇṇo yaṃ navā already requires a dissimilar vowel, so before i or ī the y never arose — which is what the verse says. The ca he reads as anuttasamuccaya, “adding what is not stated”, so that the prohibition covers iti and pati as well (itīti, patīto), where Kaccāyana’s vutti names ati alone. There is no option marker: this is a prohibition, and in its domain the general elision of 13 (§12): Sarā sare lopaṃ and the lengthening of 17 (§15): Dīghaṃ apply (atīsigaṇo). The closing lines list the three abhi phrases of the next rule and carry sare to it.
Ati iccetassa antabhūtassa ti saddassa ivaṇṇe pare ‘‘sabbo caṃ tī’’ti vuttarūpaṃ na hoti. Atissā ti ati upasaggānukaraṇametaṃ. Tenevettha vibhattilopābhāvo. Ettha ca anta saddo saddavidhinisedhappakaraṇato ati saddantabhūtaṃ ti saddameva vadati, na ivaṇṇanti daṭṭhabbaṃ, itarathā idaṃ suttameva niratthakaṃ siyā.
For the sound ti that is the end of ati, when an i-vaṇṇa follows, the form stated by 22 (§19): Sabbo canti does not arise. atissa is an imitation (anukaraṇa) of the prefix ati [quoted as a word]; that is why its inflection is not elided here. And here the word anta (“end”), since the context is one of forbidding an operation on a sound, is to be understood as denoting only the sound ti at the end of the word ati, not the i-vaṇṇa; otherwise this sutta itself would be pointless.
‘‘I vaṇṇo yaṃ navā’’tīdha, asarūpādhikārato;
Since in 21 (§21): Ivaṇṇo yaṃ navā the governance of asarūpe holds, i-vaṇṇa by ya before a homogeneous [vowel] cannot arise here [and needs no prohibition].
Ca kāro anuttasamuccayattho, tena itipatī namantassa ca na hoti. Ati isigaṇo atīsigaṇo, evaṃ atīto, atīritaṃ, itīti, itīdaṃ, patīto.
The word ca has the sense of adding what is not stated; hence [the change to c] does not take place for the end of iti and pati either. ati isigaṇo gives atīsigaṇo; likewise atīto, atīritaṃ, itīti, itīdaṃ, patīto.
Abhi akkhānaṃ, abhi uggato, abhi okāso itīdha ya kāre sampatte –
Here, in abhi akkhānaṃ, abhi uggato and abhi okāso, when [the replacement by] ya is due –
‘‘Sare’’ti vattate.
sare continues.
Buddhappiya gathers Kaccāyana’s three prefix rules §44–§46 here, directly after the ti-to-c pair, because they are of the same kind — a whole word (sabbassa) replaced before a vowel — and because abhi akkhānaṃ would otherwise fall to 21 (§21): Ivaṇṇo yaṃ navā : the lead-in at the end of 23 (§47): Atissa cantassa says so, yakāre sampatte, “when y is due”. sare is carried on from there. His gloss is one sentence and a verse. The sentence adds sabbassa, “of the whole”: abhi is replaced entire by abbha, whose final a then drops before the vowel by 13 (§12): Sarā sare lopaṃ — the pubbassaralopo of the third block, a step Kaccāyana’s vutti (abbhādeso hoti) leaves unsaid. The verse answers an objection: the sutta’s abhi is in the first form ①, yet the vutti construes it as a sixth ⑥ (abhi-iccetassa). This, he says, is required by the sense of a replacement, whose substituend is always put in the ⑥, exactly as the vutti of 149 (§82): Aṃ mo niggahītaṃ jhalapehi construes its aṃ mo. There is no option marker: outside the i-vaṇṇa case of 26 (§46): Te na vā ivaṇṇe the replacement is regular.
Abhiiccetassa sabbassa sare pare abbhā deso hoti.
For the whole of abhi, when a vowel follows, there is the replacement abbha.
‘‘Abhī’’ti paṭhamantassa, vuttiyaṃ chaṭṭhiyojanaṃ;
Though abhi [in the sutta] ends in the first form ①, its construal in the sixth ⑥ in the vutti Aṃ mo and the like.
Pubbassaralopo, abbhakkhānaṃ, abbhuggato, abbhokāso.
[Then] the elision of the preceding vowel: abbhakkhānaṃ, abbhuggato, abbhokāso.
words
replaced by §44
elided by §12
result
meaning
abhi akkhānaṃ
abbha + akkhānaṃ
aabbhakkhānaṃ slander
abhi uggato
abbha + uggato
aabbhuggato risen; spread abroad
abhi okāso
abbha + okāso
aabbhokāso the open air
abhi + akkhānaṃ
rule
→
(abhi→abbha) + akkhānaṃ
§44
→
abbh + (a) + (a + k) + khānaṃ
§10
→
abbh + (a + a) + k + khānaṃ
§12
→
(abbh + a + k) + khānaṃ
§11
→
abbhakkhānaṃ
abbhakkhānaṃ (2V 1.5.7.10: Kaṇṭakasikkhāpada)
abbhuggato (1V Veranjakanda: Verañjakaṇḍa)
abbhokāso (5V 7.5: Pañcakavāra)
Adhi agamā, adhi upagato, adhi ogāhetvā itīdha
Here, in adhi agamā, adhi upagato and adhi ogāhetvā –
The twin of 24 (§44): Abbho abhi for adhi, in Kaccāyana’s order, with sabbassa and sare carried on. Buddhappiya’s gloss is the rule with its three examples; the final a of ajjha drops before the vowel as before, by 13 (§12): Sarā sare lopaṃ . The agamā of the first example is the aorist ⏮ of gacchati with the prefix: ajjhagamā, “he attained”. The closing lines set up the next rule — abhi icchitaṃ, adhi īritaṃ, with Abbho abhi and Ajjho adhi continuing, so that 26 (§46): Te na vā ivaṇṇe can say te, “those two”. There is no option marker.
Adhi iccetassa sabbassa sare pare ajjhā deso hoti. Ajjhagamā, ajjhupagato, ajjhogāhetvā.
For the whole of adhi, when a vowel follows, there is the replacement ajjha: ajjhagamā, ajjhupagato, ajjhogāhetvā.
words
replaced by §45
elided by §12
result
meaning
adhi agamā
ajjha + agamā
aajjhagamā he attained
adhi upagato
ajjha + upagato
aajjhupagato gone over to
adhi ogāhetvā
ajjha + ogāhetvā
aajjhogāhetvā having plunged into
adhi + upagato
rule
→
(adhi→ajjha) + upagato
§45
→
ajjh + (a) + (u) + pagato
§10
→
ajjh + (a + u) + pagato
§12
→
(ajjh + u) + pagato
§11
→
ajjhupagato
ajjhagamā (12S1 8.1.2: Aratīsutta)
ajjhupagato (17A10 1.5.8: Pabbajitaabhiṇhasutta)
ajjhogāhetvā (1V 1.1.1.1: Sudinnabhāṇavāra)
Abhi icchitaṃ, adhi īritaṃ itīdha
Here, in abhi icchitaṃ and adhi īritaṃ –
‘‘Abbho abhi, ajjho adhī’’ti ca vattate.
24 (§44): Abbho abhi and 25 (§45): Ajjho adhi also continue.
Kaccāyana’s §46 in its place after §44–§45: before an i-vaṇṇa the two replacements optionally do not take place. Buddhappiya’s gloss makes the option concrete. When the replacement is withheld, the ordinary sandhi applies — elision of the i of the prefix by 13 (§12): Sarā sare lopaṃ and carrying over by 14 (§11): Naye paraṃ yutte , saralopaparanayanāni — giving abhicchitaṃ, adhīritaṃ; when it is not withheld, abbhīritaṃ, ajjhiṇamutto, ajjhiṭṭho. The vā is left unglossed and the Vā ti kiṃ? answers show the other result, so this is the open option (vikappa) of Boundaries and Domains §3.1, not a restricted one — Ruiz-Falqués does not count the rule among the forty-two vavatthitavibhāsā — though usage has in fact settled on one form for each word (abhicchitaṃ but ajjhiṭṭho). Note that the option here qualifies a prohibition: na … vā, “optionally not”. The lead-in ekaṃ idha ahaṃ prepares the next rule.
Te ca kho abhiadhi iccete upasaggā i vaṇṇe pare abbho ajjhoiti vuttarūpā na honti vā. Saralopaparanayanāni.
And those two prefixes abhi and adhi, when an i-vaṇṇa follows, optionally (vā) do not take the forms abbha and ajjha stated [above]. [Then] the elision of the vowel and the carrying-over [apply].
Abhicchitaṃ, adhīritaṃ.
abhicchitaṃ, adhīritaṃ.
words
not replaced
elided by §12
result
meaning
abhi icchitaṃ
abhi kept
iabhicchitaṃ desired
adhi īritaṃ
adhi kept
iadhīritaṃ uttered
abhi + icchitaṃ
rule
→
abh + (i) + (i + c) + chitaṃ
§10
→
abh + (i + i) + c + chitaṃ
§12
→
(abh + i + c) + chitaṃ
§11
→
abhicchitaṃ
adhi + īritaṃ
rule
→
adh + (i) + (ī) + ritaṃ
§10
→
adh + (i + ī) + ritaṃ
§12
→
(adh + ī) + ritaṃ
§11
→
adhīritaṃ
Vā ti kiṃ? Abbhīritaṃ, ajjhiṇamutto, ajjhiṭṭho.
Why vā? [Because of] abbhīritaṃ, ajjhiṇamutto, ajjhiṭṭho.
Ekamidha ahantīdha
Here, in ekaṃ idha ahaṃ –
Kaccāyana’s §20, moved from its place after §19 to follow the prefix rules: like theirs, its example is a fixed phrase. Buddhappiya reduces the rule to a single case — the dha of idha after the word eka, in the formula ekamidāhaṃ … samayaṃ — and completes the derivation with saralopadīghā: after dh has become d, the a of ida drops before ahaṃ by 13 (§12): Sarā sare lopaṃ and the surviving a is lengthened by 17 (§15): Dīghaṃ . kvaci is the exception marker: the same idha before eva keeps its dh (idheva). The sutta’s ca he uses for a change Kaccāyana’s vutti does not mention: the dha of sādhu becomes ha in sāhu dassanaṃ — before a consonant, so outside the rule’s own condition — an irregular change admitted under the ca, as 42 (§29): Vagge ghosāghosānaṃ tatiyapaṭhamā will admit others ca ggahaṇassa bahulatthattā, “because the word ca has a wide sense”. The closing lines carry navā (from 21 (§21): Ivaṇṇo yaṃ navā , over the rules between) and saramhā (from 15 (§13): Vā paro asarūpā ) into the next rule.
Dha iccetassa sare pare kvaci da kāro hoti. Eka saddato parassa idha ssa dha kārassevāyaṃ, saralopadīghā. Ekamidāhaṃ. Kvacī ti kiṃ? Idheva.
For dha, when a vowel follows, there is in some places (kvaci) da. This [applies] only to the dha of idha following the word eka; [then] the elision of the vowel and the lengthening [apply]: ekamidāhaṃ. Why kvaci? [Because of] idheva.
Ca saddena kvaci sādhu ssa dha ssa ha kāro, yathā – sāhu dassanaṃ.
By the word ca, in some places the dha of sādhu becomes ha; for example, sāhu dassanaṃ.
sāhu dassanamariyānaṃ (18Dh 15.8: Sakkavatthu)
sādhu
rule
→
sā + (dhu→hu)
§20
→
sāhu
Yathā eva tathā evātīdha
Here, in yathā eva and tathā eva –
‘‘Navā’’ti vattate, ‘‘saramhā’’ti ca.
navā continues, and saramhā.
Kaccāyana’s §22 in its place after §20 (§21 having been moved up to 21 (§21): Ivaṇṇo yaṃ navā ). navā, carried from 21 by a frog’s leap and read there as kvaci, and saramhā from 15 (§13): Vā paro asarūpā supply what the sutta leaves out, and Buddhappiya adds dīgha: the e of eva becomes ri only after a long vowel, and only after yathā and tathā — Kaccāyana’s evādissa, “of eva and the like”, he narrows to the eva that follows those two words. The two operations go together, ri for the e and shortening of the ā before it: yathariva, tathariva. navā makes them the exception, the ordinary sandhi being yatheva, tatheva by 13 (§12): Sarā sare lopaṃ ; Boundaries and Domains §4.5 lists this rule second among the six that use navā. The closing lines list the nine phrases of 30 (§70): Jhalānamiyuvā sare vā , in which the replacements y and v of 21 and 20 (§18): Vamodudantānaṃ would be due, and announce that saññā continues from the section of definitions: a designation is about to be made.
Yathātathā dvayaparassedaṃ gahaṇaṃ. Dīghasaramhā parassa eva saddādibhūtassa e kārassa ri kāro hoti, pubbo ca saro rasso hoti navā. Yathariva, tathariva. Navā ti kiṃ? Yatheva, tatheva.
This [eva] means the one that follows the two words yathā and tathā. For the e that begins the word eva following a long vowel there is ri, and the preceding vowel becomes short — navā. yathariva, tathariva. Why navā? [Because of] yatheva, tatheva.
words
e→ri and ā→a by §22result
meaning
yathā eva
yatha + riva
yathariva just as
tathā eva
tatha + riva
tathariva just so
yathā + eva
rule
→
yath + (ā) + (e) + va
§10
→
yath + ā + (e→ri) + va
§22
→
yath + (ā→a) + ri + va
§22
→
(yath + a) + ri + va
§11
→
yathariva
yathariva bhotā gotamena (6D 3.2.1: Paṭhamaibbhavāda)
tathariva (31Dht 2.7: Sattamanaya Sampayuttenavippayuttapadaniddesa)
Ti antaṃ,ti addhaṃ, aggi agāre, sattamī atthe, pañcamī antaṃ, du aṅgikaṃ, bhikkhu āsane, puthu āsane, sayambhū āsane itīdha yavā desesu sampattesu –
Here, in ti antaṃ, ti addhaṃ, aggi agāre, sattamī atthe, pañcamī antaṃ, du aṅgikaṃ, bhikkhu āsane, puthu āsane and sayambhū āsane, when the replacements ya and va [by 21 and 20] are due –
‘‘Saññā’’ti vattate.
saññā (designation) continues.
A definition from the Nāmakappa, brought into the sandhi chapter because the next rule needs it: an i-vaṇṇa is named jha and a u-vaṇṇa la. In Kaccāyana the names belong to the masculine and neuter i/u stems (aggi, bhikkhu), the feminine ones being pa by §59 (182): Te itthikhyā po , and they serve the inflection rules from §70 (30): Jhalānamiyuvā sare vā to §117 (124): Jhalato sassano vā . Buddhappiya’s gloss notes that vaṇṇa takes in the long ī and ū as well, and his verse defends the use he is about to make of the names: like pa, jha and la do not depend on being the end of a stem, since Kaccāyana himself applies them inside a verb (ākhyāte), in the middle of a stem and at the end of the first member of a compound — so ti antaṃ, du aṅgikaṃ, sattamī atthe may use 30 (§70): Jhalānamiyuvā sare vā although ti, du and sattamī are not masculine stems. saññā is carried from the Saññāvidhāna, as announced at the end of 28 (§22): Evādissa ri pubbo ca rasso . The badge is rūḷhī: the names are labels. There is no option marker.
Ivaṇṇauvaṇṇaiccete yathākkamaṃ jhala saññā honti. Vaṇṇa ggahaṇaṃ savaṇṇaggahaṇatthaṃ.
i-vaṇṇa and u-vaṇṇa respectively receive the designations jha and la. The word vaṇṇa is used so as to include the homogeneous [long] vowels.
Jhalasaññā pasaññāva, na liṅgantaṃva nissitā;
The designations jha and la, like the designation pa, do not depend on the end of a stem alone,
Kaccāyana’s §70, from the masculine i/u inflection rules of the Nāmakappa, brought here because its effect is a sandhi: before a vowel jha may become iya and la uva, so that ti antaṃ gives tiyantaṃ and bhikkhu āsane bhikkhuvāsane, where 21 (§21): Ivaṇṇo yaṃ navā and 20 (§18): Vamodudantānaṃ would give tyantaṃ and bhikkhvāsane. Buddhappiya adds the step Kaccāyana’s vutti leaves out — saralopo: the final a of iya and uva drops before the vowel by 13 (§12): Sarā sare lopaṃ . The vā he leaves unglossed, and Ruiz-Falqués (Boundaries and Domains §3.1) takes this very rule as his example of the open option (vikappa): bhikkhuvāsane and bhikkhu āsane are equally correct, and the Vā ti kiṃ? list adds the third possibility for aggi agāre, agyāgāre by 21. The closing lines set up the next rule with go ajinaṃ, go eḷakaṃ and the carry-over of go, avo and samāse: these words come from the go rules, which Buddhappiya states only in the Nāmakaṇḍa — 174 (§73): Gāva se , 170 (§75): Avaṃmhi ca and 175 (§77): Tato namaṃ patimhālutte ca samāse — so that the vattate here looks forward, not back.
Jhala iccetesaṃ iyauva iccete ādesā honti vā sare pare, saralopo.
For jha and la there are respectively the replacements iya and uva, optionally (vā), when a vowel follows; [then] the elision of the vowel.
Tiyantaṃ, tiyaddhaṃ, aggiyāgāre, sattamiyatthe, pañcamiyantaṃ, duvaṅgikaṃ, bhikkhuvāsane, puthuvāsane, sayambhuvāsane.
tiyantaṃ, tiyaddhaṃ, aggiyāgāre, sattamiyatthe, pañcamiyantaṃ, duvaṅgikaṃ, bhikkhuvāsane, puthuvāsane, sayambhuvāsane.
words
replaced by §70
elided by §12
result
meaning
ti antaṃ
tiya + antaṃ
atiyantaṃ three ends
ti addhaṃ
tiya + addhaṃ
atiyaddhaṃ three halves
aggi agāre
aggiya + agāre
aaggiyāgāre in the fire-hall
sattamī atthe
sattamiya + atthe
asattamiyatthe in the sense of the seventh form
pañcamī antaṃ
pañcamiya + antaṃ
apañcamiyantaṃ ending in the fifth form
du aṅgikaṃ
duva + aṅgikaṃ
aduvaṅgikaṃ two-factored
bhikkhu āsane
bhikkhuva + āsane
abhikkhuvāsane on the monk’s seat
puthu āsane
puthuva + āsane
aputhuvāsane on a broad seat
sayambhū āsane
sayambhuva + āsane
asayambhuvāsane on the Self-existent’s seat
ti + antaṃ
rule
→
t + (i→iya) + antaṃ
§70
→
tiy + (a) + (a + n) + taṃ
§10
→
tiy + (a + a) + n + taṃ
§12
→
(tiy + a + n) + taṃ
§11
→
tiyantaṃ
bhikkhu + āsane
rule
→
bhikkh + (u→uva) + āsane
§70
→
bhikkhuv + (a) + (ā) + sane
§10
→
bhikkhuv + (a + ā) + sane
§12
→
(bhikkhuv + ā) + sane
§11
→
bhikkhuvāsane
In sattamiyatthe, pañcamiyantaṃ and sayambhuvāsane the long ī and ū are replaced whole by iya and uva, since jha and la cover the long vowels too; in aggiyāgāre the a of agāre left standing after the elision is lengthened by 17 (§15): Dīghaṃ , as the kvaci of that rule allows, while the a of atthe and antaṃ keeps its measure.
duvaṅgikaṃ (29Dhs 2.1.2.1: Catukkanaya)
Vā ti kiṃ? Agyāgāre, sattamīatthe, bhikkhuāsane nisīdati.
Why vā? [Because of] agyāgāre, sattamīatthe, bhikkhuāsane nisīdati.
As the rule is an open option, the same words also take the other sandhi or none: aggi agāre gives agyāgāre by 21 (§21): Ivaṇṇo yaṃ navā , and sattamī atthe, bhikkhu āsane nisīdati stand unjoined.
aggi + agāre
rule
→
agg + (i) + (a) + gāre
§10
→
agg + (i→y) + a + gāre
§21
→
a + (gg→g) + y + a + gāre
§41
→
a + g + y + (a→ā) + gāre
§25
→
(a + g + y + ā) + gāre
§11
→
agyāgāre
The reduction of ggy to gy is the consonant-sandhi rule 56 (§41): Byañjano ca visaññogo , which Buddhappiya illustrates there with this very word.
agyāgāre (3V 1.12: Uruvelapāṭihāriyakathā)
Go ajinaṃ, go eḷakaṃ itīdha
Here, in go ajinaṃ and go eḷakaṃ –
‘‘Go, avo, samāse’’ti ca vattate.
go, avo and samāse also continue.
Kaccāyana’s §78, from the go rules of the Nāmakappa (§73–§81), pulled into the sandhi chapter because in gavājinaṃ, gaveḷakaṃ its effect is the sandhi of a compound: before a vowel, in a compound, the o of go becomes ava, and the final a of ava then drops by 13 (§12): Sarā sare lopaṃ , with lengthening by 17 (§15): Dīghaṃ in gavājinaṃ. The words go, avo (that is, ava) and samāse are not in this sutta; Buddhappiya announced at the end of 30 (§70): Jhalānamiyuvā sare vā that they continue, from rules he states only in the Nāmakaṇḍa. The sutta’s ca he reads with Kaccāyana’s vutti as an extension to the u-vaṇṇa, which takes uva and ava before inflections — bhuvi, the ⑦ of bhū, and pasavo, the ①⨂ of pasu — though he drops the ura of caturo from Kaccāyana’s list; those derivations belong to the Nāma and stand on the Kaccāyana page. samāse is the restriction: outside a compound go before a vowel is left alone. There is no option marker; ca is an extension. The lead-in putha eva prepares the next rule.
Go iccetassa o kārassa sare pare avā deso hoti samāse. Gavājinaṃ, gaveḷakaṃ. Ca saddaggahaṇena uvaṇṇassa uvaavā desā. Yathā – bhuvi, pasavo.
For the o of go, when a vowel follows, in a compound, there is the replacement ava: gavājinaṃ, gaveḷakaṃ. By the word ca, for a u-vaṇṇa there are the replacements uva and ava; for example, bhuvi, pasavo.
Putha evātīdha
Here, in putha eva –
Kaccāyana’s §42, brought from near the end of the chapter to stand, with 33 (§43): Pāssa canto rasso , before the great augment rule 34 (§35): Yava ma da na ta ra lā cāgamā : both insert g, the ninth augment that 34 will add by its ca. Buddhappiya’s gloss defines the term the section is about — an āgama is what “comes in”, the arising of something that was not there — and supplies the sutta’s missing ante, “at the end”: since the condition is sare, the following vowel, the augment must go next to it (nimittāsannavasena), that is, at the end of putha, which he calls a particle (nipāta). kvaci is the exception marker: puthageva beside the plain putha eva. The closing lines set up the next rule with pā eva and carry sare, go, āgamo and kvaci into it.
Putha iccetassa nipātassa ante kvaci ga kārāgamo hoti sare pare. Āgacchatīti āgamo, asantuppatti āgamo. Ettha ca ‘‘sare’’ti nimittāsannavasena puthassa anteti labbhati. Puthageva, putha eva.
At the end of the particle putha, when a vowel follows, there is in some places (kvaci) the augment (āgama) ga. An āgama is what comes in (āgacchati): the arising of something not [previously] present is an āgama. And here, from sare [being the condition], by the principle of proximity to the condition (nimittāsanna), “at the end of putha” is obtained. puthageva; [and without it] putha eva.
Pā evātīdha
Here, in pā eva –
‘‘Sare, go, āgamo, kvacī’’ti ca vattate.
sare, go, āgamo and kvaci also continue.
Kaccāyana’s §43 in its place after §42, with sare, go, āgamo and kvaci carried on: pā too takes the augment g, and by the sutta’s ca its ā is shortened at the same time — pageva, “all the more”, from pā eva. Kaccāyana’s vutti and Buddhappiya give the same example, pageva vutyassa, where vutyassa is vutti assa with the y of 21 (§21): Ivaṇṇo yaṃ navā . kvaci, carried on, marks the exception: pā eva also stands. The closing line carries vā — from 30 (§70): Jhalānamiyuvā sare vā , the nearest rule with the word — and sare into the next rule, whose sutta has neither.
Pā iccetassa ante sare pare kvaci ga kārāgamo hoti, pā ssa anto ca saro rasso hoti. Pageva vutyassa, pā eva.
At the end of pā, when a vowel follows, there is in some places (kvaci) the augment ga, and the final vowel of pā becomes short. pageva vutyassa; [and without it] pā eva.
pageva vutyassa
pā + eva
rule
→
pā + (g) + eva
§43
→
p + (ā→a) + g + eva
§43
→
pageva
‘‘Vā, sare’’ti ca vattate.
vā and sare also continue.
34 (§35): Yava ma da na ta ra lā cāgamā (And y, v, m, d, n, t, r, l insertion) apavada Section titled “34 (§35): Yava ma da na ta ra lā cāgamā (And y, v, m, d, n, t, r, l insertion) apavada ”
Kaccāyana’s §35, the augment rule, moved from the middle of the chapter to close the vowel-sandhi section after the two g rules. The sutta has no marker; Buddhappiya supplies vā from 30 (§70): Jhalānamiyuvā sare vā , as announced at the end of 33 (§43): Pāssa canto rasso , and glosses it at once: vavatthitavibhāsatthoyaṃ vā saddo, “this word vā has the sense of a restricted option” (Boundaries and Domains §2.3, §3.2). He then states the domains augment by augment. nicca: d after the prefix u (udaggo, udayo, never uaggo), and y after pari in pariyantaṃ, pariyādānaṃ, pariyesati — both marked niccaṃ. anicca: yathayidaṃ or yathā idaṃ (yathā idaṃ vā), viyañjanā or byañjanā, kenacideva or kenaci atthakāmena, sammadeva or sammā aññāya, yāvadeva or yāvāhaṃ, punadeva or punāparaṃ, chaḷabhiññā or cha abhiññā — the three Vā ti kiṃ? lists. asanta: parikkhato, upaparikkhati, where pari takes no y (idha na bhavati), and dvādhiṭṭhitaṃ, pātvākāsi, punapi, where du, pātu, puna take no r and the sandhi of 20 (§18): Vamodudantānaṃ or 13 (§12): Sarā sare lopaṃ applies instead. He also gives the reach of each augment — y after yathā, mā, taṃ, na, vi, pari, ni and before eva; v after ti, bhū, bhantā, pa; m after lahu, garu, idha, kasā, ākāse for the metre; d after u, saki, the indefinite pronouns, sammā, yāva, tāva, puna, and after ya, ta, eta, atta, sa in a compound only; n before āyati; t after yasmā, tasmā, ajja; r after ni, du, pātu, puna, dhī, pāta, catu and before iva, eva for the metre; ḷ after the numeral cha. Two further remarks: the ca adds a ninth augment, g — the puthageva, pageva of the two preceding rules and the atippago of 47 (§36): Kvaci o byañjane — where Kaccāyana’s vutti had used it for the replacements m to p, k to d and d to t; and the a for i in sakadāgāmī he puts down to the mahāvutti, the grammarians’ general licence for irregular change, which in Kaccāyana’s terms is the scope of §404 (370): Tesu vuddhilopāgamavikāraviparītādesā ca . The shortening in yathayidaṃ and sammadattho is by the consonant-sandhi rule 38 (§26): Rassaṃ .
Sare pare ya kārādayo aṭṭha āgamā honti vā. Ca saddena ga kārāgamo ca, vavatthitavibhāsatthoyaṃ vā saddo.
When a vowel follows, the eight augments beginning with ya come in, optionally (vā). By the word ca, the augment ga too. This word vā has the sense of a restricted option (vavatthitavibhāsā).
Tattha yakārāgamo yathā dibho ikārekārādīsu. Yathā idaṃ yathayidaṃ, byañjaneti adhikicca ‘‘rassa’’nti rassattaṃ, yathā idaṃ vā, yathā eva, yathāyeva, yatheva, evaṃ māyidaṃ, māyevaṃ, taṃyidaṃ, taṃyeva, nayidaṃ, nayimassa, nayimāni, navayime dhammā, buddhānaṃyeva, santiyeva, bodhiyāyeva, satiyeva, pathavīyeva, dhātuyeva, tesuyeva, soyeva, pāṭiyekkaṃ.
Among these, the augment y comes after yathā and the like, before i, e and the like. yathā idaṃ gives yathayidaṃ, with the shortening by 38 (§26): Rassaṃ under the governing word byañjane — or [unchanged] yathā idaṃ. yathā eva gives yathāyeva, or yatheva; likewise māyidaṃ, māyevaṃ, taṃyidaṃ, taṃyeva, nayidaṃ, nayimassa, nayimāni, navayime dhammā, buddhānaṃyeva, santiyeva, bodhiyāyeva, satiyeva, pathavīyeva, dhātuyeva, tesuyeva, soyeva, pāṭiyekkaṃ.
The augment y:
words
result
meaning
yathā idaṃ
yathayidaṃ (ā shortened by §26), or yathā idaṃjust as this
yathā eva
yathāyeva , or yatheva by §12just so
mā idaṃ
māyidaṃ not this
mā evaṃ
māyevaṃ not so
taṃ idaṃ
taṃyidaṃ that this
taṃ eva
taṃyeva that very
na idaṃ
nayidaṃ not this
na imassa
nayimassa not of this
na imāni
nayimāni not these
nava ime dhammā
navayime dhammā these nine things
buddhānaṃ eva
buddhānaṃyeva of the Buddhas alone
santi eva
santiyeva there are indeed
bodhiyā eva
bodhiyāyeva of awakening alone
sati eva
satiyeva mindfulness alone
pathavī eva
pathavīyeva the earth alone
dhātu eva
dhātuyeva the element alone
tesu eva
tesuyeva in those alone
so eva
soyeva that very one
pāṭi ekkaṃ
pāṭiyekkaṃ separately
yathā + idaṃ
rule
→
yathā + (y) + idaṃ
§35
→
yath + (ā→a) + y + idaṃ
§26
→
yathayidaṃ
yathayidaṃ (1V Veranjakanda: Verañjakaṇḍa)
māyidaṃ (1V 1.1.2.2: Vinītavatthu)
nayidaṃ (1V 1.1.1.1: Sudinnabhāṇavāra)
pāṭiyekkaṃ (5V 12.1: Codanādipucchāvissajjanā)
Tathā sare vipari yādito ca. Vi añjanā viyañjanā, byañjanā vā, evaṃ viyākāsi, byākāsi. Pariantaṃ pariyantaṃ, evaṃ pariyādānaṃ, pariyuṭṭhānaṃ, pariyesati, pariyosānamiti niccaṃ. Ni āyogo niyāyogo. Idha na bhavati, parikkhato, upaparikkhati.
Likewise, before a vowel, after vi, pari and the like. vi añjanā gives viyañjanā, or byañjanā [by 21]; likewise viyākāsi, byākāsi. pari antaṃ gives pariyantaṃ; likewise pariyādānaṃ, pariyuṭṭhānaṃ, pariyesati, pariyosānaṃ — always (niccaṃ). ni āyogo gives niyāyogo. Here it does not apply: parikkhato, upaparikkhati.
words
result
domain
meaning
vi añjanā
viyañjanā , or byañjanā by §21aniccathe marks
vi ākāsi
viyākāsi , or byākāsi by §21aniccahe declared
pari antaṃ
pariyantaṃ niccathe limit
pari ādānaṃ
pariyādānaṃ niccaexhaustion
pari uṭṭhānaṃ
pariyuṭṭhānaṃ niccaobsession
pari esati
pariyesati niccahe seeks
pari osānaṃ
pariyosānaṃ niccathe conclusion
ni āyogo
niyāyogo —
application
pari khato
parikkhato (no y)asantadug round
upa pari ikkhati
upaparikkhati (no y)asantahe investigates
Buddhappiya’s own vipariyādito, “after vi, pari and the like”, carries the augment it describes. The word vipari-ādito is written with the y inserted.
viyañjanā (10M 5.2: Selasutta)
Vakāro ti saddādito avaṇṇukāresu. Ti aṅgulaṃ tivaṅgulaṃ, evaṃ tivaṅgikaṃ, bhūvādayo, migī bhantā vudikkhati, pavuccati, pāguññavujutā.
The augment v comes after ti and the like, before an a-vaṇṇa and u. ti aṅgulaṃ gives tivaṅgulaṃ; likewise tivaṅgikaṃ, bhūvādayo, migī bhantā vudikkhati, pavuccati, pāguññavujutā.
words
result
meaning
ti aṅgulaṃ
tivaṅgulaṃ three finger-breadths
ti aṅgikaṃ
tivaṅgikaṃ three-factored
bhū ādayo
bhūvādayo the roots bhū and the rest
migī bhantā udikkhati
migī bhantā vudikkhati a startled doe looks up
pa uccati
pavuccati is called
pāguñña ujutā
pāguññavujutā proficiency and uprightness
migī bhantāvudikkhati (23J 18.1.2: Ummādantījātaka)
Makāro lahu ppabhutito sare chandānurakkhaṇādimhi. Lahu essati lahumessati, evaṃ garumessati, idhamāhu, kena te idha mijjhati, bhadro kasāmiva, ākāse mabhipūjayi, ekamekassa, yena midhekacce, āsatimeva.
The augment m comes after lahu and the like, before a vowel, for the sake of the metre and so on. lahu essati gives lahumessati; likewise garumessati, idhamāhu, kena te idha mijjhati, bhadro kasāmiva, ākāse mabhipūjayi, ekamekassa, yena midhekacce, āsatimeva.
words
result
meaning
lahu essati
lahumessati it will go lightly
garu essati
garumessati it will go heavily
idha āhu
idhamāhu here they said
kena te idha ijjhati
kena te idha mijjhati by what does it prosper for you here
bhadro kasā iva
bhadro kasāmiva like a thoroughbred [under] the whip
ākāse abhipūjayi
ākāse mabhipūjayi he honoured in the sky
eka ekassa
ekamekassa of each one
yena idha ekacce
yena midhekacce by which here some
āsati eva
āsatimeva it remains, just so
sittā te lahumessati (18Dh 25.7: Sambahulabhikkhuvatthu)
kena te idha mijjhati (19Vv 1.1.1: Paṭhamapīṭhavimānavatthu)
asso bhadro kasāmiva (18Dh 10.10: Pilotikatissattheravatthu)
yena midhekacce (9M 5.1: Sāleyyakasutta)
Dakāro u upasagga saki kenaci kiñci kismiñci koci sammā yāva tāva puna ya te’tattasādīhi. U upasaggato niccaṃ, u aggo udaggo, evaṃ udayo, udapādi, udāhaṭaṃ, udito, udīritaṃ, udeti.
The augment d comes after the prefix u, saki, kenaci, kiñci, kismiñci, koci, sammā, yāva, tāva, puna, ya, ta, eta, atta, sa and the like. After the prefix u always (niccaṃ): u aggo gives udaggo; likewise udayo, udapādi, udāhaṭaṃ, udito, udīritaṃ, udeti.
Nipātato ca, saki eva sakideva, evaṃ sakadāgāmi, mahāvuttisuttena i kārassa a kāro. Tathā kenacideva, kiñcideva, kismiñcideva, kocideva, sammā attho sammadattho, rassattaṃ. Evaṃ sammadakkhāto, sammadaññā vimuttānaṃ, sammadeva, yāvadatthaṃ, yāvadeva, tāvadeva, punadeva.
And after a particle: saki eva gives sakideva; likewise sakadāgāmī, where the a for the i is by the mahāvutti rule. So too kenacideva, kiñcideva, kismiñcideva, kocideva; sammā attho gives sammadattho, with shortening; likewise sammadakkhāto, sammadaññā vimuttānaṃ, sammadeva, yāvadatthaṃ, yāvadeva, tāvadeva, punadeva.
words
result
meaning
saki eva
sakideva once only
saki āgāmī
sakadāgāmī (i to a by the mahāvutti)a once-returner
kenaci eva
kenacideva by anyone at all
kiñci eva
kiñcideva anything at all
kismiñci eva
kismiñcideva in anything at all
koci eva
kocideva anyone at all
sammā attho
sammadattho (ā shortened by §26)the true goal
sammā akkhāto
sammadakkhāto well taught
sammā aññā vimuttānaṃ
sammadaññā vimuttānaṃ of those freed by right knowledge
sammā eva
sammadeva rightly indeed
yāva atthaṃ
yāvadatthaṃ as much as needed
yāva eva
yāvadeva only so far as
tāva eva
tāvadeva just so long
puna eva
punadeva once again
sammā + attho
rule
→
sammā + (d) + attho
§35
→
samm + (ā→a) + d + attho
§26
→
sammadattho
sakideva (3V 1.12: Uruvelapāṭihāriyakathā)
sammadattho vipaccati (23J 17.1.1: Tesakuṇajātaka)
sammadaññāvimuttānaṃ (18Dh 4.11: Godhikattheraparinibbānavatthu)
sammadeva (3V 1.11: Bhaddavaggiyavatthu)
yāvadeva (2V 1.5.8.2: Vilekhanasikkhāpada)
punadeva (1V 1.2.5: Sañcarittasikkhāpada)
Nāmato, yadatthaṃ, tadatthaṃ, yadantarā, tadantarā, tadaṅgavimutti, etadatthaṃ, attadatthaṃ, sadatthapasuto siyā. Yatetattasehi samāseyeva.
After a noun: yadatthaṃ, tadatthaṃ, yadantarā, tadantarā, tadaṅgavimutti, etadatthaṃ, attadatthaṃ, sadatthapasuto siyā. After ya, ta, eta, atta and sa only in a compound.
Ādi saddena aññadatthaṃ, manasādaññā vimuttānaṃ, bahudeva rattiṃ, ahudeva bhayaṃ.
By the word ādi (“and the like”): aññadatthaṃ, manasādaññā vimuttānaṃ, bahudeva rattiṃ, ahudeva bhayaṃ.
words
result
meaning
añña atthaṃ
aññadatthaṃ for another purpose
manasā aññā vimuttānaṃ
manasādaññā vimuttānaṃ of those freed by knowledge with the mind
bahu eva rattiṃ
bahudeva rattiṃ much of the night indeed
ahu eva bhayaṃ
ahudeva bhayaṃ there was just fear
bahudeva rattiṃ (3V 5.12: Mahākaccānassapañcavaraparidassana)
ahudeva bhayaṃ (6D 2.2: Komārabhaccajīvakakathā)
Vā ti kiṃ? Kenaci atthakāmena, sammā aññāya, yāvāhaṃ, tāvāhaṃ, punāparaṃ, attatthaṃ.
Why vā? [Because of] kenaci atthakāmena, sammā aññāya, yāvāhaṃ, tāvāhaṃ, punāparaṃ, attatthaṃ.
By saying vā as a restricted option, the d has an anicca domain: after the same words it is also omitted, and the phrase stays unjoined or takes the elision of 13 (§12): Sarā sare lopaṃ with the lengthening of 17 (§15): Dīghaṃ .
words
result
how
kenaci atthakāmena
kenaci atthakāmena unjoined
sammā aññāya
sammā aññāya unjoined
yāva ahaṃ
yāvāhaṃ a elided by §12, a lengthened by §15
tāva ahaṃ
tāvāhaṃ a elided by §12, a lengthened by §15
puna aparaṃ
punāparaṃ a elided by §12, a lengthened by §15
atta atthaṃ
attatthaṃ a elided by §12
punāparaṃ (19Pv 2.8: Cūḷaseṭṭhipetavatthu)
attatthaṃ (12S2 1.3.2: Dutiyadasabalasutta)
Nakāro āyatādimhi. Ito āyati ito nāyati, ciraṃ nāyati.
The augment n comes before āyati and the like: ito āyati gives ito nāyati; ciraṃ nāyati.
words
result
meaning
ito āyati
ito nāyati he comes from here
ciraṃ āyati
ciraṃ nāyati he comes after a long time
Takāro yasmā tasmā ajjādito ihaggādimhi. Yasmātiha, tasmātiha, ajjatagge.
The augment t comes after yasmā, tasmā, ajja and the like, before iha, agga and the like: yasmātiha, tasmātiha, ajjatagge.
words
result
meaning
yasmā iha
yasmātiha since … here
tasmā iha
tasmātiha therefore … here
ajja agge
ajjatagge from today onwards
tasmātiha bhikkhave (3V 1.6: Pañcavaggiyakathā)
ajjatagge (1V 1.1.1.1: Sudinnabhāṇavāra)
Rakāro ni dupātu puna dhī pāta caturādito. Ni antaraṃ nirantaraṃ, evaṃ nirālayo, nirindhano, nirīhakaṃ, niruttaro, nirojaṃ. Du atikkamo duratikkamo, durāgataṃ, duruttaṃ. Pāturahosi, pāturahesuṃ. Punarāgaccheyya, punaruttaṃ, punareva, punareti. Dhiratthu. Pātarāso.
The augment r comes after ni, du, pātu, puna, dhī, pāta, catu and the like. ni antaraṃ gives nirantaraṃ; likewise nirālayo, nirindhano, nirīhakaṃ, niruttaro, nirojaṃ. du atikkamo gives duratikkamo; durāgataṃ, duruttaṃ. pāturahosi, pāturahesuṃ. punarāgaccheyya, punaruttaṃ, punareva, punareti. dhiratthu. pātarāso.
words
result
meaning
ni antaraṃ
nirantaraṃ uninterrupted
ni ālayo
nirālayo free from attachment
ni indhano
nirindhano without fuel
ni īhakaṃ
nirīhakaṃ motionless
ni uttaro
niruttaro unsurpassed
ni ojaṃ
nirojaṃ without nourishment
du atikkamo
duratikkamo hard to overcome
du āgataṃ
durāgataṃ unwelcome
du uttaṃ
duruttaṃ ill spoken
pātu ahosi
pāturahosi it appeared
pātu ahesuṃ
pāturahesuṃ they appeared
puna āgaccheyya
punarāgaccheyya he would come again
puna uttaṃ
punaruttaṃ repeated
puna eva
punareva again indeed
puna eti
punareti he comes again
dhī atthu
dhiratthu shame on it!
pāta āso
pātarāso the morning meal
pāturahosi (3V 1.5: Brahmayācanakathā)
dhiratthu (2V 1.7.7.9: Nīcāsanasikkhāpada)
Catu saddādito, caturaṅgikaṃ, caturārakkhā, caturiddhipādapaṭilābho, caturoghanittharaṇatthaṃ. Bhatturatthe, vuttiresā, pathavīdhāturevesā.
After the word catu and the like: caturaṅgikaṃ, caturārakkhā, caturiddhipādapaṭilābho, caturoghanittharaṇatthaṃ. [And] bhatturatthe, vuttiresā, pathavīdhāturevesā.
words
result
meaning
catu aṅgikaṃ
caturaṅgikaṃ fourfold
catu ārakkhā
caturārakkhā the four protections
catu iddhipādapaṭilābho
caturiddhipādapaṭilābho attainment of the four bases of power
catu oghanittharaṇatthaṃ
caturoghanittharaṇatthaṃ for crossing the four floods
bhattu atthe
bhatturatthe in the master’s interest
vutti esā
vuttiresā this is the livelihood
pathavīdhātu eva esā
pathavīdhāturevesā this is just the earth element
bhatturatthe parakkanto (22J 3.4.1: Abbhantarajātaka)
pathavīdhāturevesā (9M 3.8: Mahāhatthipadopamasutta)
Tathā sarato ivevesu chandānurakkhaṇe. Nakkhattarājāriva tārakānaṃ, vijjurivabbhakūṭe, āraggeriva sāsapo, sāsaporiva āragge, usabhoriva, sabbhireva samāsetha.
Likewise after a vowel, before iva and eva, for the sake of the metre: nakkhattarājāriva tārakānaṃ, vijjurivabbhakūṭe, āraggeriva sāsapo, sāsaporiva āragge, usabhoriva, sabbhireva samāsetha.
words
result
meaning
nakkhattarājā iva tārakānaṃ
nakkhattarājāriva tārakānaṃ like the moon among the stars
vijju iva abbhakūṭe
vijjurivabbhakūṭe like lightning in a cloud-bank
āragge iva sāsapo
āraggeriva sāsapo like a mustard seed on an awl’s point
sāsapo iva āragge
sāsaporiva āragge like a mustard seed on an awl’s point
usabho iva
usabhoriva like a bull
sabbhi eva samāsetha
sabbhireva samāsetha one should keep company only with the good
nakkhattarājāriva tārakānaṃ (22J 11.1.2: Juṇhajātaka)
āraggeriva sāsapo (10M 5.8: Vāseṭṭhasutta)
sāsaporiva āraggā (10M 5.8: Vāseṭṭhasutta)
usabhoriva chetva bandhanāni (18Sn 1.2: Dhaniyasutta)
sabbhireva samāsetha (12S1 1.4.1: Sabbhisutta)
Vā ti kiṃ? Dvādhiṭṭhitaṃ, pātvākāsi, punapi.
Why vā? [Because of] dvādhiṭṭhitaṃ, pātvākāsi, punapi.
Lakāro chasaṅkhyāhi. Laḷānamaviseso. Cha abhiññā chaḷabhiññā, chaḷaṅgaṃ, chaḷāsīti, chaḷaṃsā, saḷāyatanaṃ.
The augment l comes after the numeral cha; there is no distinction between l and ḷ. cha abhiññā gives chaḷabhiññā; chaḷaṅgaṃ, chaḷāsīti, chaḷaṃsā, saḷāyatanaṃ.
words
result
meaning
cha abhiññā
chaḷabhiññā having the six higher knowledges
cha aṅgaṃ
chaḷaṅgaṃ sixfold
cha āsīti
chaḷāsīti eighty-six
cha aṃsā
chaḷaṃsā six-sided
cha āyatanaṃ
saḷāyatanaṃ (with sa for cha)the six sense-bases
chaḷabhiññā (12S1 8.1.7: Pavāraṇāsutta)
saḷāyatanaṃ (3V 1.1: Bodhikathā)
Vā ti kiṃ? Cha abhiññā.
Why vā? [Because of] cha abhiññā.
Iti sarasandhividhānaṃ niṭṭhitaṃ.
The section on vowel sandhi is finished.
Atha sarānameva sandhikāriye sampatte pakatibhāvo vuccate.
Now, where a sandhi operation would fall on the vowels themselves, their remaining in the original form (pakatibhāva) is stated.
‘‘Sarā, pakatī’’ti ca vattate.
And the words sarā (vowels) and pakati (the original form) are understood — supplied from the rule that follows, 36 (§23): Sarā pakati byañjane .
Buddhappiya opens his third section, the Pakatisandhividhāna (the treatment of pakati, vowels left in their original form), with this rule and puts it before 36 (§23): Sarā pakati byañjane , although Kaccāyana has the two the other way round. The vowel-sandhi rules have just ended with 34 (§35): Yava ma da na ta ra lā cāgamā , so the cases in which two vowels meet and yet nothing happens come first — sarānameva sandhikāriye sampatte pakatibhāvo vuccate — and the rule about consonants follows as the bridge into the Byañjanasandhividhāna. The words sarā and pakati that he says “continue” are therefore borrowed from the rule that follows, not carried from one that precedes.
Where the Kaccāyana vutti gives one example (ko imaṃ pathaviṃ vicessati) and one counter-example, his gloss defines pakati as the absence of elision, replacement and alteration (na lopādesavikāram āpajjante) and then lists the places where it holds: the ending of a vocative, a word-final long vowel outside a compound, the final i or u of a noun or of a past-tense verb — each with its own exception (not before iti, not where the metre must be kept, not inside a compound) — and illustrates every one of them from the canon at length. This catalogue is his; Kaccāyana’s kvaci is left unexplained.
kvaci is for him an exception marker, as Ruiz-Falqués shows from this very rule (Boundaries and Domains §4.4): it names the places where sandhi is broken off — where the metre would be broken, where the pronunciation would be awkward, and wherever the texts simply do not join the words — and in those places the elision, replacement and lengthening of 13 (§12): Sarā sare lopaṃ , 15 (§13): Vā paro asarūpā , 20 (§18): Vamodudantānaṃ , 21 (§21): Ivaṇṇo yaṃ navā and the rest do not fire. He also makes kvaci cut the other way (kvaciggahaṇena … sandhi hoti): it lets sandhi back in even at a pakati place — for a vocative before iti or for the metre, and for a word-final long vowel before iti, iva, eva, ettha and the like.
Sarā kho sare pare kvaci chandabhedāsukhuccāraṇaṭṭhāne, sandhicchārahitaṭṭhāne ca pakatirūpāni honti, na lopādesavikāramāpajjanteti attho.
Vowels, when a vowel follows, in some places keep their original form (pakatirūpa): where the metre would be broken or the pronunciation made awkward, and wherever there is no wish for sandhi (sandhicchārahitaṭṭhāna). The meaning is that they undergo no alteration by elision (lopa) or replacement (ādesa).
Tattha pakatiṭṭhānaṃ nāma ālapanantā anitismiṃ acchandānurakkhaṇe asamāse padantadīghā ca ikārukārā ca nāmapadantātītakriyādimhīti evamādi.
Here the places of the original form (pakatiṭṭhāna) are, for instance: the endings of vocatives (ālapananta), when iti does not follow and the metre does not have to be kept; word-final long vowels outside a compound; and the vowels i and u at the end of a noun (nāmapada) or of a past-tense verb (atītakriyā), and the like.
Ālapanantesu tāva – katamā cānanda aniccasaññā, katamā cānanda ādīnavasaññā, sāriputta idhekacco, ehi sivika uṭṭhehi, upāsakā idhekacco, bhoti ayye, bhikkhu arogaṃ tava sīlaṃ, siñca bhikkhu imaṃ nāvaṃ, bhikkhave evaṃ vadāmi, pañcime gahapatayo ānisaṃsā iccevamādīsu pakatibhāvo, pubbassaralopayavā desādayo na honti.
First, at the ending of a vocative: in the following and the like the original form remains; the elision of the preceding vowel, the replacement by y or v and the other operations do not take place. The vocative is the word before the + in each note.
katamā cānanda aniccasaññā (no sandhi for ānanda + aniccasaññā) (17A10 2.1.10: Girimānandasutta)
katamā cānanda ādīnavasaññā (no sandhi for ānanda + ādīnavasaññā) (17A10 2.1.10: Girimānandasutta)
sāriputta idhekacco (no sandhi for sāriputta + idhekacco) (17A9 1.2.2: Saupādisesasutta)
ehi sivika uṭṭhehi (no sandhi for sivika + uṭṭhehi)
upāsakā idhekacco (no sandhi for upāsakā + idhekacco)
bhoti ayye (no sandhi for bhoti + ayye)
bhikkhu arogaṃ tava sīlaṃ (no sandhi for bhikkhu + arogaṃ)
siñca bhikkhu imaṃ nāvaṃ (no sandhi for bhikkhu + imaṃ) (18Dh 25.7: Sambahulabhikkhuvatthu)
bhikkhave evaṃ vadāmi (no sandhi for bhikkhave + evaṃ)
pañcime gahapatayo ānisaṃsā (no sandhi for gahapatayo + ānisaṃsā; pañca + ime does join) (3V 6.14: Pāṭaligāmavatthu)
Kvaci ggahaṇena iti smiṃ chandānurakkhaṇe sandhi hoti, yathā – sakkā devīti, namo te buddha vīratthu.
Sare ti kiṃ? Sādhu mahārājāti, evaṃ kira bhikkhūti.
By saying sare (before a vowel) the vocative keeps its form only before a vowel. Before a consonant it is open to the rules of the next section: in sādhu mahārājāti and evaṃ kira bhikkhūti the i of iti has gone by 15 (§13): Vā paro asarūpā , and the vocative, now standing before the consonant t, is lengthened by 37 (§25): Dīghaṃ .
sādhu mahārājāti (mahārāja + iti)
evaṃ kira bhikkhūti (bhikkhu + iti)
mahārāja + iti
rule
→
mahārāja + (i)ti
§13
→
mahārāj(a→ā) + ti
§25
→
mahārājāti
Asamāse padantadīghesu – āyasmā ānando gāthā ajjhabhāsi, devā ābhassarā yathā, tevijjā iddhippattā ca, bhagavā uṭṭhāyāsanā, bhagavā etadavoca, abhivādetvā ekamantaṃ aṭṭhāsi, gantvā olokento, bhūtavādī atthavādī, yaṃ itthī arahā assa, sāmāvatī āha, pāpakārī ubhayattha tappati, nadī ottharati, ye te bhikkhū appicchā, bhikkhū āmantesi, bhikkhū ujjhāyiṃsu, bhikkhū evamāhaṃsu, imasmiṃ gāme ārakkhakā, sabbe ime, katame ekādasa, gambhīre odakantike, appamādo amatapadaṃ, saṅgho āgacchatu, ko imaṃ pathaviṃ viccessati, āloko udapādi, eko ekāya, cattāro oghā.
For word-final long vowels outside a compound — ā, ī, ū, e and o in turn — the following. In each the long vowel before the + is left as it is.
āyasmā ānando gāthā ajjhabhāsi (no sandhi for āyasmā + ānando, gāthā + ajjhabhāsi)
devā ābhassarā yathā (no sandhi for devā + ābhassarā) (12S1 4.2.8: Piṇḍasutta)
tevijjā iddhippattā ca (no sandhi for tevijjā + iddhippattā; the Tipiṭaka spells iddhipattā) (12S1 6.1.5: Aññatarabrahmasutta)
bhagavā uṭṭhāyāsanā (no sandhi for bhagavā + uṭṭhāya; uṭṭhāya + āsanā does join) (1V 1.2.8: Duṭṭhadosasikkhāpada)
bhagavā etadavoca (no sandhi for bhagavā + etad) (1V Veranjakanda: Verañjakaṇḍa)
abhivādetvā ekamantaṃ aṭṭhāsi (no sandhi for abhivādetvā + ekamantaṃ) (1V 1.2.8: Duṭṭhadosasikkhāpada)
gantvā olokento (no sandhi for gantvā + olokento)
bhūtavādī atthavādī (no sandhi for bhūtavādī + atthavādī) (6D 1.2.1: Cūḷasīla)
yaṃ itthī arahā assa (no sandhi for itthī + arahā)
sāmāvatī āha (no sandhi for sāmāvatī + āha)
pāpakārī ubhayattha tappati (no sandhi for pāpakārī + ubhayattha) (18Dh 1.12: Devadattavatthu)
nadī ottharati (no sandhi for nadī + ottharati)
ye te bhikkhū appicchā (no sandhi for bhikkhū + appicchā) (1V 1.1.2: Dutiyapārājikasikkhāpada)
bhikkhū āmantesi (no sandhi for bhikkhū + āmantesi) (1V 1.1.1.1: Sudinnabhāṇavāra)
bhikkhū ujjhāyiṃsu (no sandhi for bhikkhū + ujjhāyiṃsu)
bhikkhū evamāhaṃsu (no sandhi for bhikkhū + evaṃ) (1V 1.1.2: Dutiyapārājikasikkhāpada)
imasmiṃ gāme ārakkhakā (no sandhi for gāme + ārakkhakā)
sabbe ime (no sandhi for sabbe + ime)
katame ekādasa (no sandhi for katame + ekādasa) (17A11 1.2.5: Mettāsutta)
gambhīre odakantike (no sandhi for gambhīre + odakantike) (18Kh 8: Nidhikaṇḍasutta)
appamādo amatapadaṃ (no sandhi for appamādo + amatapadaṃ) (18Dh 2.1: Sāmāvatīvatthu)
saṅgho āgacchatu (no sandhi for saṅgho + āgacchatu)
ko imaṃ pathaviṃ viccessati (no sandhi for ko + imaṃ; the Tipiṭaka spells vicessati) (18Dh 4.1: Pathavikathāpasutapañcasatabhikkhuvatthu)
āloko udapādi (no sandhi for āloko + udapādi) (3V 1.6: Pañcavaggiyakathā)
eko ekāya (no sandhi for eko + ekāya) (1V 1.3.1: Paṭhamaaniyatasikkhāpada)
cattāro oghā (no sandhi for cattāro + oghā) (8D 10.4: Catukka)
Nipātesupi – are ahampi, sace imassa kāyassa, no abhikkamo, aho acchariyo, atho anto ca, atha kho āyasmā, atho oṭṭhavacittakā, tato āmantayī satthā.
Kvaci ggahaṇena a kāra itīvevetthādīsu sandhipi. Yathā – āgatattha, āgatamhā, katamāssu cattāro, appassutāyaṃ puriso, itthīti, ca marīva, sabbeva, sveva, eseva nayo, parisuddhetthāyasmanto, nettha, taṃ kutettha labbhā, sacesa brāhmaṇa, tathūpamaṃ, yathāha.
By the word kvaci there is sandhi as well [after a word-final long vowel] when the vowel a follows and before iti, iva, eva, ettha and the like, as in the following. The first four join a long final with a following a; the next six show iti, iva, eva and ettha; the last three are “the like”.
lemma
from
inflected
rule
meaning
🚹①⨂(āgata) ▶️🤘⨂(atthi)
āgatā attha
āgatattha
§12
you have come
🚹①⨂(āgata) ▶️👆⨂(atthi)
āgatā amhā
āgatamhā
§12
we have come
🚹①⨂(katama) 🔼(assu) 🚹①⨂(catu)
katamā assu cattāro
katamāssu cattāro
§12, §15
which then are the four?
🚹①⨀(appassuta) 🚹①⨀(ima) 🚹①⨀(purisa)
appassuto ayaṃ puriso
appassutāyaṃ puriso
§12, §15
this man is of little learning
🚺①⨀(itthī) 🔼(iti)
itthī iti
itthīti
§13
“a woman”
🚺①⨀(camarī) 🔼(iva)
camarī iva
camarīva
§13
like a yak-cow
🚹①⨂(sabba) 🔼(eva)
sabbe eva
sabbeva
§12
all indeed
🚹①⨀(ta) 🔼(eva)
so eva
sveva
§18
he himself
🚹①⨀(eta) 🔼(eva) 🚹①⨀(naya)
eso eva nayo
eseva nayo
§12
this very method
🚹①⨂(parisuddha) 🔼(ettha) 🚹①⨂(āyasmantu)
parisuddhā ettha āyasmanto
parisuddhetthāyasmanto
§12
the venerable ones are pure in this
🔼(na) 🔼(ettha)
na ettha
nettha
§12
not here
🚻①⨀(ta) 🔼(kuto) 🔼(ettha) 🔼(labbhā)
taṃ kuto ettha labbhā
taṃ kutettha labbhā
§12
how could that be got here?
🔼(sace) 🚹①⨀(eta) 🚹⓪⨀(brāhmaṇa)
sace eso brāhmaṇa
sacesa brāhmaṇa
§12, §27
if this, brahmin
🔼(tathā) 🚺②⨀(upamā)
tathā upamaṃ
tathūpamaṃ
§12, §15
a simile to that effect
🔼(yathā) ⏮🤟⨀(āha)
yathā āha
yathāha
§12
as he said
Citations for the phrases the Tipiṭaka has in this shape:
appassutāyaṃ puriso (18Dh 11.7: Lāḷudāyītheravatthu)
parisuddhetthāyasmanto (1V 1.1.4.2: Vinītavatthu)
taṃ kutettha labbhā (1V Veranjakanda: Verañjakaṇḍa)
Asamāse ti kiṃ? Jivhāyatanaṃ, avijjogho, itthindriyaṃ, abhibhāyatanaṃ, bhayatupaṭṭhānaṃ. Acchandānurakkhaṇe ti kiṃ? Saddhīdha vittaṃ purisassa seṭṭhaṃ, yo missaro.
By saying asamāse (outside a compound) the word-final long vowel is left alone only between separate words; inside a compound it joins with the vowel that follows.
lemma
from
inflected
rule
meaning
🚻①⨀(jivhāyatana)
jivhā + āyatanaṃ
jivhāyatanaṃ
§12
the tongue base
🚹①⨀(avijjogha)
avijjā + ogho
avijjogho
§12
the flood of ignorance
🚻①⨀(itthindriya)
itthī + indriyaṃ
itthindriyaṃ
§12
the faculty of femininity
🚻①⨀(abhibhāyatana)
abhibhū + āyatanaṃ
abhibhāyatanaṃ
§12
a base of mastery
🚻①⨀(bhayatupaṭṭhāna)
bhayato + upaṭṭhānaṃ
bhayatupaṭṭhānaṃ
§12
the appearance as fearful
By saying acchandānurakkhaṇe (where the metre need not be kept) the long final stands only where the verse does not require otherwise; where it does, sandhi is made even here — the elision with lengthening of saddhā idha and the m inserted by 34 (§35): Yava ma da na ta ra lā cāgamā in yo issaro.
saddhīdha vittaṃ purisassa seṭṭhaṃ (saddhā + idha) (12S1 1.8.3: Vittasutta)
yo missaro (yo + issaro) (23J 22.1.9.9: Kāḷāgirikaṇḍa)
saddhā + idha
rule
→
saddh(ā) + idha
§12
→
saddh + (i→ī)dha
§15
→
saddhīdha
Nāmapadantaikārukāresu – gāthāhi ajjhabhāsi, pupphāni āhariṃsu, satthu adāsi.
For i and u at the end of a noun: the following.
gāthāhi ajjhabhāsi (no sandhi for gāthāhi + ajjhabhāsi) (1V 1.2.6: Kuṭikārasikkhāpada)
pupphāni āhariṃsu (no sandhi for pupphāni + āhariṃsu)
satthu adāsi (no sandhi for satthu + adāsi)
The past-tense verb of the list above (nāmapadantātītakriyādimhi) is illustrated by adāsi ahaṃ and udapādi in the phrases already given: their final i too stands before a vowel.
Kvacī ti kiṃ? Manasākāsi.
Kaccāyana states this rule first (§23) and Sare kvaci after it; Buddhappiya reverses the pair and closes his short Pakatisandhividhāna with this one, so that its word byañjane can at once be declared the governing word (adhikāra) of the whole Byañjanasandhividhāna that follows — ‘‘Byañjane’’ti adhikāro — while sarā and the kvaci of 35 (§24): Sare kvaci run on into 37–39. His gloss adds a single word to the Kaccāyana vutti, yebhuyyena, “for the most part”: leaving a vowel unaltered before a consonant — no lengthening, no shortening, no elision — is the general state of affairs, and the three rules that follow, 37 (§25): Dīghaṃ , 38 (§26): Rassaṃ and 39 (§27): Lopañca tatrākāro , each marked kvaci, are its exceptions. Ruiz-Falqués (Boundaries and Domains §2.2) counts this yebhuyyena among the words by which Buddhappiya controls optionality beside the four proper markers. The sutta itself has no option marker; it is the general rule (utsarga) for vowels before consonants, as Kaccāyana’s page has it.
Sarā kho byañjane pare pakatirūpā hontīti yebhuyyena dīgharassalopehi vikārābhāvo. Yathā – accayo, paccayo, bhāsati vā karoti vā, vedanākkhandho, bhāgyavā, bhadro kasāmiva, dīyati, tuṇhībhūto, so dhammaṃ deseti.
Vowels, when a consonant follows, keep their original form: that is, for the most part (yebhuyyena) there is no alteration by lengthening, shortening or elision. For example the following.
lemma
inflected
the vowel left alone
meaning
🚹①⨀(accaya)
accayo
a before ccpassing, transgression
🚹①⨀(paccaya)
paccayo
a before cccondition
▶️🤟⨀(bhāsati) 🔼(vā) ▶️🤟⨀(karoti) 🔼(vā)
bhāsati vā karoti vā
i before v, ā before kspeaks or acts
🚹①⨀(vedanākkhandha)
vedanākkhandho
ā before kkhthe aggregate of feeling
🚹①⨀(bhāgyavantu)
bhāgyavā
ā before gyfortunate
🚹①⨀(bhadra) 🚺③⨀(kasā) 🔼(iva)
bhadro kasāmiva
o before k, ā before mlike a good [horse touched] by the whip
🔵▶️🤟⨀(dīyati)
dīyati
ī before yis given
🚹①⨀(tuṇhībhūta)
tuṇhībhūto
ī before bhfallen silent
🚹①⨀(ta) 🚹②⨀(dhamma) ▶️🤟⨀(deseti)
so dhammaṃ deseti
o before dh, aṃ before dhe teaches the Dhamma
bhāsati vā karoti vā (18Dh 1.1: Cakkhupālattheravatthu)
bhadro kasāmiva (18Dh 10.10: Pilotikatissattheravatthu)
Iti pakatisandhividhānaṃ niṭṭhitaṃ.
Thus ends the treatment of pakatisandhi, the vowels that keep their original form.
Atha byañjanasandhi vuccate.
Now consonant sandhi (byañjanasandhi) is stated.
‘‘Byañjane’’ti adhikāro, ‘‘sarā, kvacī’’ti ca vattate.
byañjane (before a consonant), from 36 (§23): Sarā pakati byañjane , is the governing word (adhikāra) of what follows, and sarā (vowels) and kvaci (in some places) continue.
The first exception to 36 (§23): Sarā pakati byañjane . Kaccāyana too puts §25 straight after §23–§24, and Buddhappiya keeps the run Dīghaṃ – Rassaṃ – Lopañca tatrākāro (37–39), the three alterations a vowel may suffer before a consonant, before turning to what happens to the consonant itself (40–42). The sutta is the single word dīghaṃ; sarā, byañjane and kvaci are all supplied by the governing word and the continuations he has just declared. Where the Kaccāyana vutti has only kvaci, his gloss names the three places in which the lengthening occurs — in a canonical passage, for ease of pronunciation and to keep the metre — and his examples are sorted by them: tyāssa, svāssa, svākkhāto, yvāhaṃ, where a y or v replacement by 19 (§17): Yamedantassādeso or 20 (§18): Vamodudantānaṃ is followed by lengthening; the verse lines munī care, khantī paramaṃ, maṅkū bhavissāmi, jāyatī soko, ujū ca suhujū ca; and the prefixes anu, du, su lengthened before p, r and h (anūpaghāto, dūrakkhaṃ, sūrakkhaṃ, dūharatā). kvaci is an exception marker (Boundaries and Domains §4.2): it isolates these cases from the general rule 36, and where it does not reach — tyajja, svassa, patiliyyati — the vowel stays short. The long and the short forms are not an open option but assigned to different words.
Sarā kho byañjane pare kvaci dīghaṃ pappontīti suttasukhuccāraṇachandānurakkhaṇaṭṭhānesu dīgho.
Vowels, when a consonant follows, in some places become long: the lengthening takes place in a canonical passage, for ease of pronunciation, and in order to keep the metre.
Tyassa pahīnā tyāssa pahīnā, svassa svāssa, madhuva maññati bālo madhuvā maññatī bālo, tathā evaṃ gāme munī care, khantī paramaṃ tapo titikkhā, na maṅkū bhavissāmi, svākkhāto, yvāhaṃ, kāmato jāyatī soko, kāmato jāyatī bhayaṃ, sakko ujū ca suhujū ca, anūpaghāto, dūrakkhaṃ, dūramaṃ, sūrakkhaṃ, dūharatā.
The vowel that is lengthened is marked in the derivation column; where another rule has first replaced a vowel by y or v or elided one, it is named before §25.
lemma
derivation
rule
meaning
🚹①⨂(ta) 🚹⑥⨀(ta) 🚹①⨂(pahīna)
te + assa → ty(a→ā)ssa → tyāssa pahīnā
§17, §25
those are abandoned for him
🚹①⨀(ta) 🚹⑥⨀(ta)
so + assa → sv(a→ā)ssa → svāssa
§18, §25
he, for him
🚻①⨀(madhu) 🔼(iva) ▶️🤟⨀(maññati) 🚹①⨀(bāla)
madhu + iva → madhuv(a→ā); maññat(i→ī) → madhuvā maññatī bālo
§13, §25
the fool thinks it is like honey
🔼(evaṃ) 🚹⑦⨀(gāma) 🚹①⨀(muni) ⏯🤟⨀(carati)
mun(i→ī) → evaṃ gāme munī care
§25
so should the sage go about the village
🚺①⨀(khanti) 🚻①⨀(parama) 🚻①⨀(tapa) 🚺①⨀(titikkhā)
khant(i→ī) → khantī paramaṃ tapo titikkhā
§25
patience, forbearance, is the highest austerity
🔼(na) 🚹①⨀(maṅku) ⏯👆⨀(bhavati)
maṅk(u→ū) → na maṅkū bhavissāmi
§25
I shall not be downcast
🚹①⨀(svākkhāta)
su + akkhāto → sv(a→ā)kkhāto → svākkhāto
§18, §25
well proclaimed
🚹①⨀(ya) ⚧️①⨀(amha)
yo + ahaṃ → yv(a→ā)haṃ → yvāhaṃ
§18, §25
I who
🚹⑤⨀(kāma) ▶️🤟⨀(jāyati) 🚹①⨀(soka)
jāyat(i→ī) → kāmato jāyatī soko
§25
from sensual desire grief is born
🚹⑤⨀(kāma) ▶️🤟⨀(jāyati) 🚻①⨀(bhaya)
jāyat(i→ī) → kāmato jāyatī bhayaṃ
§25
from sensual desire fear is born
🚹①⨀(sakka) 🚹①⨀(uju) 🔼(ca) 🚹①⨀(suhuju) 🔼(ca)
uj(u→ū), suhuj(u→ū) → sakko ujū ca suhujū ca
§25
able, upright and truly upright
🚹①⨀(anūpaghāta)
anu + upaghāto → an(u→ū)paghāto → anūpaghāto
§12, §25
not harming
🚻①⨀(dūrakkha)
du + rakkhaṃ → d(u→ū)rakkhaṃ → dūrakkhaṃ
§25
hard to guard
🚻①⨀(dūrama)
du + ramaṃ → d(u→ū)ramaṃ → dūramaṃ
§25
hard to take delight in
🚻①⨀(sūrakkha)
su + rakkhaṃ → s(u→ū)rakkhaṃ → sūrakkhaṃ
§25
well guarded
🚺①⨀(dūharatā)
du + haratā → d(u→ū)haratā → dūharatā
§25
the state of being hard to remove
The verse lines as the Tipiṭaka has them:
tyāssa pahīnā (11M 4.10: Dhātuvibhaṅgasutta)
madhuṃvā maññati bālo (the Tipiṭaka’s reading of Buddhappiya’s madhuvā maññatī bālo) (18Dh 5.10: Uppalavaṇṇattherīvatthu)
evaṃ gāme munī care (18Dh 4.5: Macchariyakosiyaseṭṭhivatthu)
khantī paramaṃ tapo titikkhā (7D 1.16: Cārikāanujānana)
yvāhaṃ (3V 10.2: Dīghāvuvatthu)
kāmato jāyatī soko, kāmato jāyatī bhayaṃ (18Dh 16.5: Anitthigandhakumāravatthu)
sakko ujū ca suhujū ca (18Kh 9: Mettasutta)
anūpaghāto (7D 1.16: Cārikāanujānana)
Kvacī ti kiṃ? Tyajja, svassa, patiliyyati.
Yiṭṭhaṃ vā hutaṃ vā loke, yadi vā sāvake, puggalā dhammadasā te, bhovādī nāma so hoti, yathābhāvī guṇena so itīdha –
In yiṭṭhaṃ vā hutaṃ vā loke, yadi vā sāvake, puggalā dhammadasā te, bhovādī nāma so hoti, yathābhāvī guṇena so — here,
Pubbasmiṃyevādhikāre –
under the same governing words as before [sarā, byañjane, kvaci] —
The second exception to 36 (§23): Sarā pakati byañjane , placed as Kaccāyana places it, directly after 37 (§25): Dīghaṃ . The words sarā, byañjane and kvaci are supplied once more — pubbasmiṃyevādhikāre, “under the same governing words as before” — and the lead-in gives the unshortened phrases first so that the rule can be watched at work on them. Kaccāyana’s vutti has two examples and three counter-examples; Buddhappiya’s gloss sorts the shortening into three places and illustrates each: before a consonant to keep the metre (yiṭṭhaṃva, puggala, bhovādi), before an inserted consonant (āgama — the y and d of 34 (§35): Yava ma da na ta ra lā cāgamā in yathayidaṃ, sammadakkhāto) and before a conjunct (saṃyoga: parakkamo, assādo, taṇhakkhayo, lābhimhi, vasimhi, thullaccayo). His counter-examples answer the same three heads in the same order: no shortening for the metre in māyidaṃ, none before the inserted d of manasādaññā, none before the conjuncts of yathākkamaṃ, ākhyātikaṃ, dīyyati, sūyyati. kvaci again isolates the exception (Boundaries and Domains §4.2): the shortened and the unshortened forms are assigned to different words, not left to choice.
Sarā kho byañjane pare kvaci rassaṃ pappontīti chandānurakkhaṇe, āgame, saṃyoge ca rassattaṃ.
Vowels, when a consonant follows, in some places become short: the shortening takes place to keep the metre, before an inserted letter (āgama), and before a conjunct (saṃyoga).
Chandānurakkhaṇe tāva yiṭṭhaṃva hutaṃva loke, yadiva sāvake, puggala dhammadasā te, bhovādi nāma so hoti, yathābhāvi guṇena so.
Āgame yathayidaṃ, sammadakkhāto.
Saṃyoge parākamo parakkamo, āsādo assādo, evaṃ taṇhakkhayo, jhānassa lābhimhi, vasimhi, thullaccayo.
Kvacī ti kiṃ? Māyidaṃ, manasā daññā vimuttānaṃ, yathākkamaṃ, ākhyātikaṃ, dīyyati, sūyyati.
By saying kvaci (in some places) the long vowel is kept in other cases of the same three kinds: before the inserted y of māyidaṃ and the inserted d of manasādaññā, and before the conjuncts of yathākkamaṃ, ākhyātikaṃ, dīyyati and sūyyati.
lemma
inflected
the vowel kept
meaning
🔼(mā) 🚻①⨀(ima)
māyidaṃ
ā before inserted ynot this
🚹③⨀(mana) 🔽(aññāya) 🚹⑥⨂(vimutta)
manasā daññā vimuttānaṃ
ā before inserted dof those released through understanding by the mind
🔼(yathākkamaṃ)
yathākkamaṃ
ā before kkin due order
🚻①⨀(ākhyātika)
ākhyātikaṃ
ā before khythe verb
🔵▶️🤟⨀(dīyati)
dīyyati
ī before yyis given
🔵▶️🤟⨀(sūyati)
sūyyati
ū before yyis heard
māyidaṃ (1V 1.1.2.2: Vinītavatthu)
Eso kho byantiṃ kāhiti, so gacchaṃ na nivattati iccatra –
In eso kho byantiṃ kāhiti and so gacchaṃ na nivattati — here,
Tasmiṃyevādhikāre –
under the same governing words —
The third and last alteration of a vowel before a consonant; Kaccāyana places it in the same way (§27 after §26). Buddhappiya narrows the sutta’s general sarā … lopaṃ to the one case the examples cover — etatasaddanta-okārass' evāyaṃ lopo, “this elision is of the final o of the words eta and ta alone” — an explicit limit where the Kaccāyana vutti only shows examples. He uses the ca of the sutta to extend the same treatment to a following vowel (esa attho, esa ābhogo, esa idāni), where otherwise 13 (§12): Sarā sare lopaṃ would have dropped the o and put nothing in its place. The rule’s two operations are welded together: the o goes and an a comes into the vacated place (tatra lutte ṭhāne akārāgamo), and this is what later lets him split the sutta (yogavibhāga) and use tatrākāro on its own to insert an a where no vowel was elided — see 59 (§51): Anupadiṭṭhānaṃ vuttayogato . The kvaci here does not exclude: eso dhammo and esa dhammo, so muni and sa muni are both good Pāli, and Ruiz-Falqués counts this rule among the three in which Buddhappiya’s kvaci in effect marks an open option, a vikappa (Boundaries and Domains §4.1). sarā, byañjane and kvaci are supplied as before (tasmiṃyevādhikāre).
Sarā kho byañjane pare kvaci lopaṃ papponti, tatra lutte ṭhāne a kārāgamo ca hoti. Etata saddantokārassevāyaṃ lopo.
Vowels, when a consonant follows, in some places are elided, and in that place, where [the vowel] has been elided, the letter a is inserted. This elision is only of the final o of the words eta and ta.
Esa kho byantiṃ kāhiti, sa gacchaṃ na nivattati, evaṃ esa dhammo, esa pattosi, sa muni, sa sīlavā.
lemma
derivation
rule
meaning
🚹①⨀(eta) 🔼(kho) 🚻②⨀(byanti) ⏯🤟⨀(karoti)
eso + kho → es(o) + kho → es + (a) + kho → esa kho byantiṃ kāhiti
§27
this one will make an end
🚹①⨀(ta) 🚹①⨀(gacchanta) 🔼(na) ▶️🤟⨀(nivattati)
so + gacchaṃ → s(o) + gacchaṃ → s + (a) + gacchaṃ → sa gacchaṃ na nivattati
§27
he, going, does not turn back
🚹①⨀(eta) 🚹①⨀(dhamma)
eso + dhammo → esa dhammo
§27
this is the nature of things
🚹①⨀(eta) 🚹①⨀(patta) ▶️🤘⨀(atthi)
eso + pattosi → esa pattosi
§27
this you have attained
🚹①⨀(ta) 🚹①⨀(muni)
so + muni → sa muni
§27
that sage
🚹①⨀(ta) 🚹①⨀(sīlavantu)
so + sīlavā → sa sīlavā
§27
he is virtuous
esa kho byanti kāhiti (the Tipiṭaka spells byanti) (18Dh 24.7: Cūḷadhanuggahapaṇḍitavatthu)
esa pattosi (18Dh 10.4: Koṇḍadhānattheravatthu)
sa sīlavā (18Dh 6.9: Dhammikattheravatthu)
Kvacī ti kiṃ? Eso dhammo, so muni, so sīlavā.
Ca saddena eta saddantassa sarepi kvaci lopo. Yathā – esa attho, esa ābhogo, esa idāni.
Vipariṇāmena ‘‘saramhā, byañjanassā’’ti ca vattate.
With a change of inflection form (vipariṇāma) the words continue [into the next rule] as saramhā (after a vowel) and byañjanassa (of a consonant).
From the vowel Buddhappiya now turns to the consonant that follows it: 40–42 are the consonant doublings, and he has just changed the inflection forms of the carried words to fit — saramhā (after a vowel) and byañjanassa (of a consonant), where 37–39 had sarā … byañjane. Kaccāyana’s §28 stands at the same point of his own sequence. The Kaccāyana vutti gives six examples and one counter-example and leaves ṭhāne (“in the [proper] place”) to the reader; Buddhappiya’s gloss is a small treatise on it. He first defines dvebhāva (dvinnaṃ bhāvo, “being two”, doubling), then lists the places: after a short a, the initial consonant of the prefixes pa, pati, paṭi and of the roots kamu, kusa, kudha, kī, gaha, juta, ñā, si, su, sambhū, sara, sasa and the like; the initial of tika, taya, tiṃsa, vata and the like; the final of the roots vatu, vaṭu, disa and the like; the consonant after the prefixes u, du, ni, after the pronoun ta, the numerals catu and cha, and the sa that stands for santa; and y and the like after a long vowel other than ā inside a word — to which a memory verse adds the replacements of 41 (§269): Yavataṃ talanadakārānaṃ byañjanāni calañajakārattaṃ , a replacement made together with a root-final, and the s of the suffix ssī after tapa — and finally the doubling made for the metre and the aspirates that 42 (§29): Vagge ghosāghosānaṃ tatiyapaṭhamā will govern. Each head is then illustrated in turn, and each has its own Ṭhāne ti kiṃ? answer, so that the one word ṭhāne does the work of an option marker: Ruiz-Falqués lists ṭhāne here and at 42 as an equivalent of kvaci (Boundaries and Domains §2.2). The doubling is not open to choice but confined to the places named, and mā ca pamādo, nikāyo, sūyati show where it stops. The sutta has no marker of its own; the vā he needs for the next rule is fetched at the end by a frog’s leap from 277 (§260): Sabbassa so dāmhi vā (Kaccāyana §260 (277): Sabbassa so dāmhi vā ).
Saramhā parassa byañjanassa dvebhāvo hoti ṭhāne. Dvinnaṃ bhāvo dvibhāvo, so eva dvebhāvo.
After a vowel the consonant that follows is doubled, in the [proper] place. dvibhāva is “being two” (dvinnaṃ bhāvo), and dvebhāva is the same word.
Ettha ca ṭhānaṃ nāma rassākārato paraṃ pa pati paṭikamukusa kudha kī gaha juta ñāsi su sambhū sara sasādīnamādibyañjanānaṃ dvebhāvaṃ, tika taya tiṃsa vatādīnamādi ca, vatu vaṭudisādīnamantañca, u du ni-upasagga ta catu cha santasaddādesādiparañca, apadantā nākāradīghato ya kārādi ca,
And here the “place” (ṭhāna) is:
after a short a, the doubling of the initial consonant of [the prefixes] pa, pati, paṭi and of [the roots] kamu, kusa, kudha, kī, gaha, juta, ñā, si, su, sambhū, sara, sasa and the like;
the initial of tika, taya, tiṃsa, vata and the like;
the final of [the roots] vatu, vaṭu, disa and the like;
what follows the prefixes u, du, ni, [the pronoun] ta, [the numerals] catu, cha, the replacement of the word santa [sa] and the like;
and y and the like after a long vowel other than ā that is not at the end of a word;
Yavataṃ talanādīna-mādeso ca sayādinaṃ;
and the replacement of t, l, n and the rest that carry a y; the replacement, together with the root-final, in [forms with] ya and the like; s of [the suffix] ssī after tapa and the like;
Chandānurakkhaṇe ca – ghara jhe dhaṃsu bhamādīnamādi ca, rassākārato vaggānaṃ catutthadutiyā ca iccevamādi.
and [doubling] to keep the metre; and the initial of [the roots] ghara, jhe, dhaṃsu, bhamu and the like, and the fourth and second letters of the classes after a short a — and so on.
Tattha pa pati paṭīsu tāva – idha pamādo idhappamādo, evaṃ appamādo, vippayutto, suppasanno, sammā padhānaṃ sammappadhānaṃ, rassattaṃ, appativattiyo, adhipatippaccayo, suppatiṭṭhito, appaṭipuggalo, vippaṭisāro, suppaṭipanno, suppaṭipatti.
Of these, first for pa, pati and paṭi: idha pamādo becomes idhappamādo, and in the same way the rest. In sammappadhānaṃ there is also shortening (38 (§26): Rassaṃ ).
lemma
derivation
rule
meaning
🔼(idha) 🚹①⨀(pamāda)
idha + pamādo → idha + (p→pp)amādo → idhappamādo
§28
negligence here
🚹①⨀(appamāda)
a + pamādo → a + (p→pp)amādo → appamādo
§28
heedfulness
🚹①⨀(vippayutta)
vi + payutto → vippayutto
§28
disjoined
🚹①⨀(suppasanna)
su + pasanno → suppasanno
§28
well pleased
🚻①⨀(sammappadhāna)
sammā + padhānaṃ → samm(ā→a) + (p→pp)adhānaṃ → sammappadhānaṃ
§28, §26
right endeavour
🚹①⨀(appativattiya)
a + pativattiyo → appativattiyo
§28
not to be turned back
🚹①⨀(adhipatippaccaya)
adhipati + paccayo → adhipatippaccayo
§28
the predominance condition
🚹①⨀(suppatiṭṭhita)
su + patiṭṭhito → suppatiṭṭhito
§28
well established
🚹①⨀(appaṭipuggala)
a + paṭipuggalo → appaṭipuggalo
§28
without a peer
🚹①⨀(vippaṭisāra)
vi + paṭisāro → vippaṭisāro
§28
remorse
🚹①⨀(suppaṭipanna)
su + paṭipanno → suppaṭipanno
§28
practising well
🚺①⨀(suppaṭipatti)
su + paṭipatti → suppaṭipatti
§28
good practice
Kamādidhātūsu – pakkamo, paṭikkamo, hetukkamo, ākamati akkamati, evaṃ parakkamati, yathākkamaṃ.
Pakkosati, paṭikkosati, anukkosati, ākosati, akkosati.
kusa:
lemma
derivation
rule
meaning
▶️🤟⨀(pakkosati)
pa + kosati → pakkosati
§28
calls, summons
▶️🤟⨀(paṭikkosati)
paṭi + kosati → paṭikkosati
§28
objects to
▶️🤟⨀(anukkosati)
anu + kosati → anukkosati
§28
calls after
▶️🤟⨀(akkosati)
ā + kosati → (ā→a) + (k→kk)osati → akkosati
§28, §26
abuses
Akkuddho, atikkodho.
kudha:
lemma
derivation
rule
meaning
🚹①⨀(akkuddha)
a + kuddho → akkuddho
§28
not angered
🚹①⨀(atikkodha)
ati + kodho → atikkodho
§28
excessive anger
Dhanakkīto, vikkayo, anukkayo.
kī:
lemma
derivation
rule
meaning
🚹①⨀(dhanakkīta)
dhana + kīto → dhanakkīto
§28
bought with money
🚹①⨀(vikkaya)
vi + kayo → vikkayo
§28
sale
🚹①⨀(anukkaya)
anu + kayo → anukkayo
§28
buying in turn
Paggaho, viggaho, anuggaho, niggaho, candaggaho, diṭṭhiggāho.
gaha:
lemma
derivation
rule
meaning
🚹①⨀(paggaha)
pa + gaho → paggaho
§28
exertion
🚹①⨀(viggaha)
vi + gaho → viggaho
§28
quarrel
🚹①⨀(anuggaha)
anu + gaho → anuggaho
§28
assistance
🚹①⨀(niggaha)
ni + gaho → niggaho
§28
restraint
🚹①⨀(candaggaha)
canda + gaho → candaggaho
§28
eclipse of the moon
🚹①⨀(diṭṭhiggāha)
diṭṭhi + gāho → diṭṭhiggāho
§28
clinging to a view
Pajjoto, vijjotati, ujjoto.
juta:
lemma
derivation
rule
meaning
🚹①⨀(pajjota)
pa + joto → pajjoto
§28
a lamp
▶️🤟⨀(vijjotati)
vi + jotati → vijjotati
§28
shines forth
🚹①⨀(ujjota)
u + joto → ujjoto
§28
radiance
Kataññū, viññū, paññāṇaṃ, viññāṇaṃ, anuññā, manuññā, samaññā.
ñā:
lemma
derivation
rule
meaning
🚹①⨀(kataññū)
kata + ñū → kataññū
§28
grateful
🚹①⨀(viññū)
vi + ñū → viññū
§28
discerning
🚻①⨀(paññāṇa)
pa + ñāṇaṃ → paññāṇaṃ
§28
a mark, a sign
🚻①⨀(viññāṇa)
vi + ñāṇaṃ → viññāṇaṃ
§28
consciousness
🚺①⨀(anuññā)
anu + ñā → anuññā
§28
permission
🚺①⨀(manuññā)
manu + ñā → manuññā
§28
pleasing
🚺①⨀(samaññā)
sama + ñā → samaññā
§28
designation
Avassayo, nissayo, samussayo.
si:
lemma
derivation
rule
meaning
🚹①⨀(avassaya)
ava + sayo → avassayo
§28
support
🚹①⨀(nissaya)
ni + sayo → nissayo
§28
reliance
🚹①⨀(samussaya)
samu + sayo → samussayo
§28
accumulation; the body
Appassuto, vissuto, bahussuto, āsavā assavā.
su:
lemma
derivation
rule
meaning
🚹①⨀(appassuta)
appa + suto → appassuto
§28
of little learning
🚹①⨀(vissuta)
vi + suto → vissuto
§28
renowned
🚹①⨀(bahussuta)
bahu + suto → bahussuto
§28
very learned
🚹①⨂(assava)
ā + savā → (ā→a) + (s→ss)avā → assavā
§28, §26
attentive, obedient
Passambhanto, vissambhati.
sambhū:
lemma
derivation
rule
meaning
🚹①⨀(passambhanta)
pa + sambhanto → passambhanto
§28
becoming calm
▶️🤟⨀(vissambhati)
vi + sambhati → vissambhati
§28
is confident
Aṭṭassaro, vissarati, anussarati.
sara:
lemma
derivation
rule
meaning
🚹①⨀(aṭṭassara)
aṭṭa + saro → aṭṭassaro
§28
a cry of distress
▶️🤟⨀(vissarati)
vi + sarati → vissarati
§28
forgets
▶️🤟⨀(anussarati)
anu + sarati → anussarati
§28
recollects
Passasanto, vissasanto, mahussasanto, āsāso assāso.
sasa:
lemma
derivation
rule
meaning
🚹①⨀(passasanta)
pa + sasanto → passasanto
§28
breathing out
🚹①⨀(vissasanta)
vi + sasanto → vissasanto
§28
trusting
🚹①⨀(mahussasanta)
mahu + sasanto → mahussasanto
§28
sighing deeply
🚹①⨀(assāsa)
ā + sāso → (ā→a) + (s→ss)āso → assāso
§28, §26
breathing in; comfort
Avissajjento, vissajjento, pariccajanto, upaddavo, upakkiliṭṭho, mittaddu, āyabbayo, udabbahi iccādi.
“And the like” covers roots outside the list:
lemma
derivation
rule
meaning
🚹①⨀(avissajjenta)
avi + sajjento → avissajjento
§28
not letting go
🚹①⨀(vissajjenta)
vi + sajjento → vissajjento
§28
giving away; answering
🚹①⨀(pariccajanta)
pari + cajanto → pariccajanto
§28
giving up
🚹①⨀(upaddava)
upa + davo → upaddavo
§28
misfortune
🚹①⨀(upakkiliṭṭha)
upa + kiliṭṭho → upakkiliṭṭho
§28
defiled
🚹①⨀(mittaddu)
mitta + du → mittaddu
§28
a betrayer of friends
🚹①⨀(āyabbaya)
āya + vayo → āya + (v→bb)ayo → āyabbayo
§28
income and expenditure
⏮🤟⨀(udabbahati)
uda + vahi → uda + (v→bb)ahi → udabbahi
§28
pulled out
The v of vaya and vahi becomes bb, not vv: the doubled v is written bb, as 42 (§29): Vagge ghosāghosānaṃ tatiyapaṭhamā will note under bo vassa.
Saramhā ti kiṃ? Sampayutto, sampatijāto, sampaṭicchannaṃ, saṅkamanto, saṅgaho.
Ṭhāne ti kiṃ? Mā ca pamādo, patigayhati, vacīpakopaṃ rakkheyya, ye pamattā yathā matā, manopakopaṃ rakkheyya, idha modati, pecca modati.
Tikādīsu – kusalattikaṃ, pītittikaṃ, hetuttikaṃ, vedanāttikaṃ, lokattayaṃ, bodhittayaṃ, vatthuttayaṃ. Ekattiṃsa, dvattiṃsa, catuttiṃsa. Sīlabbataṃ, subbato, sappītiko, samannāgato, punappunaṃ iccādi.
For tika and the rest — the initial of tika, taya, tiṃsa, vata doubled after the preceding word, and “the like”:
lemma
derivation
rule
meaning
🚻①⨀(kusalattika)
kusala + tikaṃ → kusala + (t→tt)ikaṃ → kusalattikaṃ
§28
the triad of the wholesome
🚻①⨀(pītittika)
pīti + tikaṃ → pītittikaṃ
§28
the triad of rapture
🚻①⨀(hetuttika)
hetu + tikaṃ → hetuttikaṃ
§28
the triad of roots
🚻①⨀(vedanāttika)
vedanā + tikaṃ → vedanāttikaṃ
§28
the triad of feeling
🚻①⨀(lokattaya)
loka + tayaṃ → lokattayaṃ
§28
the three worlds
🚻①⨀(bodhittaya)
bodhi + tayaṃ → bodhittayaṃ
§28
the three awakenings
🚻①⨀(vatthuttaya)
vatthu + tayaṃ → vatthuttayaṃ
§28
the three objects [of refuge]
⚧️①⨀(ekattiṃsa)
eka + tiṃsa → ekattiṃsa
§28
thirty-one
⚧️①⨀(dvattiṃsa)
dva + tiṃsa → dvattiṃsa
§28
thirty-two
⚧️①⨀(catuttiṃsa)
catu + tiṃsa → catuttiṃsa
§28
thirty-four
🚻①⨀(sīlabbata)
sīla + vataṃ → sīla + (v→bb)ataṃ → sīlabbataṃ
§28
precepts and observances
🚹①⨀(subbata)
su + vato → subbato
§28
of good conduct
🚹①⨀(sappītika)
sa + pītiko → sappītiko
§28
accompanied by rapture
🚹①⨀(samannāgata)
samanu + āgato → saman(u) + (n→nn)āgato → samannāgato
§12, §28
endowed with
🔼(punappunaṃ)
puna + punaṃ → punappunaṃ
§28
again and again
Vatu vaṭu disādīnamante yathā – vattati, vaṭṭati, dassanaṃ, phasso iccādi.
U du ni upasaggādiparesu – ukaṃso ukkaṃso. Dukaraṃ dukkaraṃ, nikaṅkho nikkaṅkho3
After the prefixes u, du, ni and the rest: ukaṃso becomes ukkaṃso, dukaraṃ dukkaraṃ, nikaṅkho nikkaṅkho.
lemma
derivation
rule
meaning
🚹①⨀(ukkaṃsa)
u + kaṃso → u + (k→kk)aṃso → ukkaṃso
§28
excellence
🚻①⨀(dukkara)
du + karaṃ → dukkaraṃ
§28
hard to do
🚹①⨀(nikkaṅkha)
ni + kaṅkho → nikkaṅkho
§28
free from doubt
Evaṃ uggataṃ, duccaritaṃ, nijjaṭaṃ, uññātaṃ, unnati, uttaro, dukkaro, niddaro, unnato, duppañño, dubbalo, nimmalo, uyyutto, dullabho, nibbatto, ussāho, dussaho, nissāro.
In the same way:
lemma
derivation
rule
meaning
🚻①⨀(uggata)
u + gataṃ → uggataṃ
§28
risen
🚻①⨀(duccarita)
du + caritaṃ → duccaritaṃ
§28
misconduct
🚻①⨀(nijjaṭa)
ni + jaṭaṃ → nijjaṭaṃ
§28
disentangled
🚻①⨀(uññāta)
u + ñātaṃ → uññātaṃ
§28
despised
🚺①⨀(unnati)
u + nati → unnati
§28
elevation
🚹①⨀(uttara)
u + taro → uttaro
§28
higher
🚹①⨀(dukkara)
du + karo → dukkaro
§28
difficult
🚹①⨀(niddara)
ni + daro → niddaro
§28
free from anguish
🚹①⨀(unnata)
u + nato → unnato
§28
raised
🚹①⨀(duppañña)
du + pañño → duppañño
§28
foolish
🚹①⨀(dubbala)
du + balo → dubbalo
§28
weak
🚹①⨀(nimmala)
ni + malo → nimmalo
§28
stainless
🚹①⨀(uyyutta)
u + yutto → uyyutto
§28
on the march
🚹①⨀(dullabha)
du + labho → dullabho
§28
hard to obtain
🚹①⨀(nibbatta)
ni + vatto → ni + (v→bb)atto → nibbatto
§28
arisen
🚹①⨀(ussāha)
u + sāho → ussāho
§28
effort
🚹①⨀(dussaha)
du + saho → dussaho
§28
hard to bear
🚹①⨀(nissāra)
ni + sāro → nissāro
§28
without substance
Tathā takkaro, tajjo, tanninno, tappabhavo, tammayo.
Catukkaṃ, catuddisaṃ, catuppādo, catubbidhaṃ, catussālaṃ.
After catu:
lemma
derivation
rule
meaning
🚻①⨀(catukka)
catu + kaṃ → catukkaṃ
§28
a set of four
🚻①⨀(catuddisa)
catu + disaṃ → catuddisaṃ
§28
the four quarters
🚹①⨀(catuppāda)
catu + pādo → catuppādo
§28
four-footed
🚻①⨀(catubbidha)
catu + vidhaṃ → catu + (v→bb)idhaṃ → catubbidhaṃ
§28
fourfold
🚻①⨀(catussāla)
catu + sālaṃ → catussālaṃ
§28
a building with four halls
Chakkaṃ, channavuti, chappadikā, chabbassāni.
After cha:
lemma
derivation
rule
meaning
🚻①⨀(chakka)
cha + kaṃ → chakkaṃ
§28
a set of six
🚺①⨀(channavuti)
cha + navuti → channavuti
§28
ninety-six
🚺①⨀(chappadikā)
cha + padikā → chappadikā
§28
six-footed
🚻①⨂(chabbassa)
cha + vassāni → cha + (v→bb)assāni → chabbassāni
§28
six years
Sakkāro, sakkato, saddiṭṭhi, sappuriso, mahabbalo.
Ṭhāne ti kiṃ? Nikāyo, nidānaṃ, nivāso, nivāto, tato, catuvīsati, chasaṭṭhi.
Yakārādimhi – nīyyati, sūyyati, abhibhūyya, viceyya, vineyya, dheyyaṃ, neyyaṃ, seyyo, jeyyo, veyyākaraṇo.
Ādi saddena etto, ettāvatā.
Anākāra ggahaṇaṃ kiṃ? Mālāya, dolāya, samādāya.
Ṭhāne ti kiṃ? Upanīyati, sūyati, toyaṃ.
Yavatamādese – jāti andho, vipali āso, ani āyo, yadi evaṃ, api ekacce, api ekadā iccatra, i kārassa ‘‘ivaṇṇo yaṃ navā’’ti ya kāre kate –
For the replacement of the consonants that carry a y: in jāti andho, vipali āso, ani āyo, yadi evaṃ, api ekacce, api ekadā — here, when the i has been made y by 21 (§21): Ivaṇṇo yaṃ navā —
‘‘Sabbassa so dāmhi vā’’ti ito maṇḍūkagatiyā vā ti vattate.
the word vā continues by a frog’s leap (maṇḍūkagati) from 277 (§260): Sabbassa so dāmhi vā .
41 (§269): Yavataṃ talanadakārānaṃ byañjanāni calañajakārattaṃ (t, l, ṇ, d with y become c, l, ñ, j) apavada Section titled “41 (§269): Yavataṃ talanadakārānaṃ byañjanāni calañajakārattaṃ (t, l, ṇ, d with y become c, l, ñ, j) apavada ”
Kaccāyana keeps this rule for the very end of the Nāmakappa (§269), where the suffix ṇya of §261 (369): Avaṇṇo ye lopañca leaves a y against the final consonant of the base (bāhussaccaṃ, kāruññaṃ). Buddhappiya lifts it into the consonant sandhi, between the general doubling and the aspirate rule, because the same y arises across a word boundary whenever 21 (§21): Ivaṇṇo yaṃ navā turns a final i into y before a vowel — jāti andho, jātyandho, jaccandho — and because its result is a doubled consonant, the business of 40 (§28): Paradvebhāvo ṭhāne and 42 (§29): Vagge ghosāghosānaṃ tatiyapaṭhamā . His gloss adds three things to the Kaccāyana vutti. First the purpose of the word kāra, which the Kaccāyana vutti uses only for ma: it covers the consonant that carries the y in general, so that a y after s or after a letter of the k, c, ṭ and p classes takes the form of that letter (appekacce), a y after t, dh, ṇ gives ch, jh, ñ, and the replacement is then doubled by 40; and the memory verse of 40 had already announced two further heads illustrated here, the ssī of tapassī, yasassī and doubling for the metre. Second, a vā: the Kaccāyana sutta has none, and its page reads the rule as obligatory, but Buddhappiya brings one by a frog’s leap from 277 (§260): Sabbassa so dāmhi vā and gives forms in which the change is not made, paṭisanthāravutyassa, bālyaṃ, ālasyaṃ. He does not call it a vavatthitavibhāsā; his Vā ti kiṃ? presents an open option (vikappa, Boundaries and Domains §3.1) between the assimilated and the unassimilated conjunct, though each word has in practice settled for one of them. Third, since for him the rule stands under the saramhā carried from 40, a counter-example for it: in añāyo (ñāya from nyāya) and ākāsānañcāyatanaṃ (ānañca from ananta + ya) the consonant that carried the y did not stand after a vowel, and the result is a single ñ, c, not a doubled one.
Ya kāravantānaṃ talanada kārānaṃ saṃyogabyañjanāni yathākkamaṃ calañaja kārattamāpajjante vā.
The conjunct consonants t, l, n and d that carry a y become, in that order, c, l, ñ and j — optionally.
Kāra ggahaṇaṃ ya vataṃ sa kāra ka ca ṭa pa vaggānaṃ sa kārakacaṭapa vaggādesatthaṃ, tathā ya vataṃ ta dha ṇa kārānaṃ cha jhaña kārādesatthañca, tato yavatamādesassa anena dvibhāvo.
The word kāra is included so that [the y of] an s, or of a letter of the k, c, ṭ or p class, that carries a y is replaced by that s or class letter, and likewise so that [the y of] a t, dh or ṇ that carries a y is replaced by ch, jh or ñ; and then the replacement of the consonant that carried the y is doubled by the preceding rule, 40 (§28): Paradvebhāvo ṭhāne .
Jaccandho, vipallāso, aññāyo, yajjevaṃ, appekacce appekadā.
Vā ti kiṃ? Paṭisanthāravutyassa, bālyaṃ, ālasyaṃ.
Saramhā ti kiṃ? Añāyo, ākāsānañcāyatanaṃ.
Tapā dito si mhi – tapassī, yasassī.
Chandānurakkhaṇe – nappajjahe vaṇṇabalaṃ purāṇaṃ, ujjugatesu seyyo, gacchanti suggatinti.
To keep the metre:
lemma
derivation
rule
meaning
🔼(na) ⏯🤟⨀(pajahati) 🚻②⨀(vaṇṇabala) 🚻②⨀(purāṇa)
na + pajahe → na + (p→pp)a(j→jj)ahe → nappajjahe vaṇṇabalaṃ purāṇaṃ
§28
he would not lose his old complexion and strength
🚹⑦⨂(ujjugata) 🚹①⨀(seyya)
ujugatesu → u(j→jj)ugatesu → ujjugatesu seyyo
§28
better [than a gift] to the upright
▶️🤟⨂(gacchati) 🚺②⨀(suggati) 🔼(iti)
sugatiṃ → su(g→gg)atiṃ → gacchanti suggatinti
§28
“they go to a good destination”
nappajjahe vaṇṇabalaṃ purāṇaṃ (22J 4.1.3: Ekarājajātaka)
ujjugatesu seyyo (18Dh 8.7: Sāriputtatherassa sahāyakabrāhmaṇavatthu)
gacchanti suggatiṃ (12S1 2.3.1: Sivasutta)
Vaggacatutthadutiyesu pana tasmiṃyevādhikāre ‘‘paradvebhāvo ṭhāne’’ti ca vattate.
But for the fourth and second letters of the classes, under the same governing words, [the rule] 40 (§28): Paradvebhāvo ṭhāne also continues.
Vaggacatutthadutiyānaṃ sadisavasena dvibhāve sampatte niyamatthamāha.
When the doubling of the fourth and second letters of the classes by way of a like letter would apply, he states [the next rule] as a restriction (niyama).
The rule completes the doubling. Kaccāyana’s §29 follows §28 in the same way, and Buddhappiya’s lead-in says how he reads it: as a niyama, a restriction on 40 (§28): Paradvebhāvo ṭhāne . Where an aspirate — the fourth or second letter of a class — would be doubled “by way of a like letter” (gh + gh), the doubling takes the shape of the third or first letter of that class instead (ggh, kkh), the letter nearest to it in place of articulation (ṭhānāsannavasena). His gloss then argues that ghosāghosānaṃ must mean the fourth and second letters only, although “voiced” would take in the nasals and y r l v h too: otherwise kataññū, tannayo, tammayo would come out with a third letter, an unwanted consequence (aniṭṭhappasaṅga) — or one may say the governing ṭhāne excludes them. The examples follow by head — the aspirates after pa, u, du, ni; the y-replacements bojjhaṅgā, bujjhati of 41 (§269): Yavataṃ talanadakārānaṃ byañjanāni calañajakārattaṃ ; the second letters kh, ch, ṭh, th, ph after a short a, class by class; and after ā, which the list of 40 had left out — each head with its own Saramhā ti kiṃ? and Ṭhāne ti kiṃ?. ṭhāne is again Buddhappiya’s kvaci (Boundaries and Domains §2.2, on Rūp 40 and 42): nidhānaṃ, mahādhanaṃ, kammaphalaṃ are places where the aspirate is left single.
The last third of the article is his own. The ca of 27 (§20): Do dhassa ca , he says, has the sense of bahula, “variously”, and under it he gathers the consonant substitutions that no Kaccāyana rule states: the aspiration of k, p, c, t after ni (nikkhamati, nicchayo), b for v (kubbanto, nibbānaṃ), l for r (palibodho, taluno), ṭ for t (dukkaṭaṃ), k for t (niyako), y for j (niyaṃputtaṃ), k for g (kulūpako) and ṇ for n (paṇidhānaṃ) — each kvaci, for the words in which usage has it, where the Kaccāyana vutti to §20 had used the ca for the single case sāhu.
Saramhā parabhūtānaṃ vagge ghosāghosānaṃ byañjanānaṃ yathākkamaṃ vaggatatiya paṭhamakkharā dvibhāvaṃ gacchanti ṭhāneti ṭhānāsannavasena tabbagge tatiyapaṭhamāva honti.
Of the voiced and unvoiced consonants of a class that stand after a vowel, the third and the first letters of the class respectively undergo the doubling, in the [proper] place — that is, by nearness of place [of articulation] it is the third and first letters of that same class that appear.
Ettha ca sampatte niyamattā ghosāghosa ggahaṇena catutthadutiyāva adhippetā, itarathā aniṭṭhappasaṅgosiyā. Tena ‘‘kataññū, tannayo, tammayo’’tiādīsu vaggapañcamānaṃ satipi ghosatte tatiyappasaṅgona hoti, ṭhānādhikārato vā.
And here, since the rule is a restriction (niyama) on a doubling that would already apply, by the words ghosa and aghosa only the fourth and second letters are meant; otherwise an unwanted consequence would follow. Hence in kataññū, tannayo, tammayo and the like, although the fifth letters of the classes are voiced, there is no occasion for a third letter — or else [they are excluded] by the governing word ṭhāne.
Gharādīsu pa u du niādiparacatutthesu tāva – pagharati paggharati, evaṃ uggharati, ugghāṭeti, dugghoso, nigghoso, eseva cajjhānaphalo, paṭhamajjhānaṃ, abhijjhāyati, ujjhāyati, viddhaṃseti, uddhaṃsito, uddhāro, niddhāro, niddhano, niddhuto, vibbhanto, ubbhato, dubbhikkhaṃ, nibbhayaṃ, tabbhāvo, catuddhā, catubbhi4
First, for ghara and the other roots, and for the fourth letters after pa, u, du, ni and the rest: pagharati becomes paggharati, and in the same way the others. The doubling letter is the third of the class: ggh, jjh, ddh, bbh.
lemma
derivation
rule
meaning
▶️🤟⨀(paggharati)
pa + gharati → pa + (gh→ggh)arati → paggharati
§29
trickles
▶️🤟⨀(uggharati)
u + gharati → uggharati
§29
oozes out
▶️🤟⨀(ugghāṭeti)
u + ghāṭeti → ugghāṭeti
§29
unfastens
🚹①⨀(dugghosa)
du + ghoso → dugghoso
§29
an ill sound
🚹①⨀(nigghosa)
ni + ghoso → nigghoso
§29
an utterance
🚹①⨀(eta) 🔼(eva) 🔼(ca) 🚹①⨀(jhānaphala)
ca + jhānaphalo → ca + (jh→jjh)ānaphalo → eseva cajjhānaphalo
§29
and this is the fruit of jhāna
🚻①⨀(paṭhamajjhāna)
paṭhama + jhānaṃ → paṭhamajjhānaṃ
§29
the first jhāna
▶️🤟⨀(abhijjhāyati)
abhi + jhāyati → abhijjhāyati
§29
covets
▶️🤟⨀(ujjhāyati)
u + jhāyati → ujjhāyati
§29
complains
▶️🤟⨀(viddhaṃseti)
vi + dhaṃseti → vi + (dh→ddh)aṃseti → viddhaṃseti
§29
destroys
🚹①⨀(uddhaṃsita)
u + dhaṃsito → uddhaṃsito
§29
discredited
🚹①⨀(uddhāra)
u + dhāro → uddhāro
§29
removal
🚹①⨀(niddhāra)
ni + dhāro → niddhāro
§29
specification
🚹①⨀(niddhana)
ni + dhano → niddhano
§29
penniless
🚹①⨀(niddhuta)
ni + dhuto → niddhuto
§29
shaken off
🚹①⨀(vibbhanta)
vi + bhanto → vi + (bh→bbh)anto → vibbhanto
§29
gone astray
🚹①⨀(ubbhata)
u + bhato → ubbhato
§29
withdrawn
🚻①⨀(dubbhikkha)
du + bhikkhaṃ → dubbhikkhaṃ
§29
famine
🚻①⨀(nibbhaya)
ni + bhayaṃ → nibbhayaṃ
§29
fearless
🚹①⨀(tabbhāva)
ta + bhāvo → tabbhāvo
§29
that state
🔼(catuddhā)
catu + dhā → catuddhā
§29
in four ways
⚧️③⨂(catu)
catu + bhi → catubbhi
§29
by four
🔼(chaddhā)
cha + dhā → chaddhā
§29
in six ways
🚹①⨀(saddhamma)
sa + dhammo → saddhammo
§29
the true Dhamma
🚹①⨀(sabbhūta)
sa + bhūto → sabbhūto
§29
real
🚹①⨀(mahaddhana)
mahā + dhano → mah(ā→a) + (dh→ddh)ano → mahaddhano
§29, §26
very rich
🚻①⨀(mahabbhaya)
mahā + bhayaṃ → mah(ā→a) + (bh→bbh)ayaṃ → mahabbhayaṃ
§29, §26
great danger
Yavata mādesādīsu – bojjhaṅgā, āsabbhaṃ, bujjhitabbaṃ, bujjhati.
Ṭhāne ti kiṃ? Sīlavantassa jhāyino, ye jhānappasutā dhīrā, nidhānaṃ, mahādhanaṃ.
Rassākāratoparaṃ vaggadutiyesu – pañca khandhā pañcakkhandhā, evaṃrūpakkhandho, akkhamo, abhikkhaṇaṃ, avikkhepo, jātikkhettaṃ, dhātukkhobho, āyukkhayo. Setachattaṃ setacchattaṃ, evaṃ sabbacchannaṃ, vicchinnaṃ, bodhicchāyā, jambucchāyā, samucchedo. Tatra ṭhito tatraṭṭhito, evaṃ thalaṭṭhaṃ, jalaṭṭhaṃ, adhiṭṭhitaṃ, niṭṭhitaṃ, cattāriṭṭhānāni, garuṭṭhāniyo, samuṭṭhito, suppaṭṭhāno. Yasatthero, yattha, tattha, pattharati, vitthāro, abhitthuto, vitthambhito, anutthunaṃ, catuttho, kuttha. Papphoṭeti, mahapphalaṃ, nipphalaṃ, vipphāro, paripphuseyya, madhupphāṇitaṃ.
For the second letters after a short a [and the other short vowels], class by class — the doubling letter is the first of the class: kkh, cch, ṭṭh, tth, pph. pañca khandhā becomes pañcakkhandhā, setachattaṃ setacchattaṃ, tatra ṭhito tatraṭṭhito, and in the same way the rest.
lemma
derivation
rule
meaning
🚹①⨂(pañcakkhandha)
pañca + khandhā → pañca + (kh→kkh)andhā → pañcakkhandhā
§29
the five aggregates
🚹①⨀(rūpakkhandha)
rūpa + khandho → rūpakkhandho
§29
the aggregate of form
🚹①⨀(akkhama)
a + khamo → akkhamo
§29
impatient
🔼(abhikkhaṇaṃ)
abhi + khaṇaṃ → abhikkhaṇaṃ
§29
constantly
🚹①⨀(avikkhepa)
avi + khepo → avikkhepo
§29
non-distraction
🚻①⨀(jātikkhetta)
jāti + khettaṃ → jātikkhettaṃ
§29
the field of birth
🚹①⨀(dhātukkhobha)
dhātu + khobho → dhātukkhobho
§29
disturbance of the elements
🚹①⨀(āyukkhaya)
āyu + khayo → āyukkhayo
§29
the end of life
🚻①⨀(setacchatta)
seta + chattaṃ → seta + (ch→cch)attaṃ → setacchattaṃ
§29
the white parasol
🚻①⨀(sabbacchanna)
sabba + channaṃ → sabbacchannaṃ
§29
completely covered
🚻①⨀(vicchinna)
vi + chinnaṃ → vicchinnaṃ
§29
cut off
🚺①⨀(bodhicchāyā)
bodhi + chāyā → bodhicchāyā
§29
the shade of the Bodhi tree
🚺①⨀(jambucchāyā)
jambu + chāyā → jambucchāyā
§29
the shade of a rose-apple tree
🚹①⨀(samuccheda)
samu + chedo → samucchedo
§29
cutting off
🚹①⨀(tatraṭṭhita)
tatra + ṭhito → tatra + (ṭh→ṭṭh)ito → tatraṭṭhito
§29
standing there
🚻①⨀(thalaṭṭha)
thala + ṭhaṃ → thalaṭṭhaṃ
§29
on dry land
🚻①⨀(jalaṭṭha)
jala + ṭhaṃ → jalaṭṭhaṃ
§29
in the water
🚻①⨀(adhiṭṭhita)
adhi + ṭhitaṃ → adhiṭṭhitaṃ
§29
resolved upon
🚻①⨀(niṭṭhita)
ni + ṭhitaṃ → niṭṭhitaṃ
§29
finished
🚻①⨂(catu) 🚻①⨂(ṭhāna)
cattāri + ṭhānāni → cattāriṭṭhānāni
§29
the four states
🚹①⨀(garuṭṭhāniya)
garu + ṭhāniyo → garuṭṭhāniyo
§29
one in a teacher’s place
🚹①⨀(samuṭṭhita)
samu + ṭhito → samuṭṭhito
§29
arisen
🚹①⨀(suppaṭṭhāna)
suppa + ṭhāno → suppaṭṭhāno
§29
well established
🚹①⨀(yasatthera)
yasa + thero → yasa + (th→tth)ero → yasatthero
§29
the Elder Yasa
🔼(yattha)
ya + tha → yattha
§29
where
🔼(tattha)
ta + tha → tattha
§29
there
▶️🤟⨀(pattharati)
pa + tharati → pattharati
§29
spreads out
🚹①⨀(vitthāra)
vi + thāro → vitthāro
§29
extent, detail
🚹①⨀(abhitthuta)
abhi + thuto → abhitthuto
§29
praised
🚹①⨀(vitthambhita)
vi + thambhito → vitthambhito
§29
stiffened
🚹①⨀(anutthunanta)
anu + thunaṃ → anutthunaṃ
§29
lamenting
🚹①⨀(catuttha)
catu + tho → catuttho
§29
fourth
🔼(kuttha)
ku + tha → kuttha
§29
where
▶️🤟⨀(papphoṭeti)
pa + phoṭeti → pa + (ph→pph)oṭeti → papphoṭeti
§29
shakes off
🚻①⨀(mahapphala)
maha + phalaṃ → mahapphalaṃ
§29
of great fruit
🚻①⨀(nipphala)
ni + phalaṃ → nipphalaṃ
§29
fruitless
🚹①⨀(vipphāra)
vi + phāro → vipphāro
§29
pervasion
⏯🤟⨀(paripphusati)
pari + phuseyya → paripphuseyya
§29
would touch all over
🚻①⨀(madhupphāṇita)
madhu + phāṇitaṃ → madhupphāṇitaṃ
§29
honey and molasses
cattāri ṭhānāni naro pamatto (the Tipiṭaka prints the words apart) (18Dh 22.4: Khemakaseṭṭhiputtavatthu)
Ākārato – ākhāto ākkhāto, evaṃ taṇhākkhayo, āṇākkhettaṃ, saññākkhandho, āchādayati, ācchādayati, evaṃ ācchindati, nāvāṭṭhaṃ, āttharati, āpphoṭeti.
Saramhā ti kiṃ? Saṅkhāro, taṅkhaṇe, sañchannaṃ, taṇṭhānaṃ, santhuto, tamphalaṃ.
Ṭhāne ti kiṃ? Pūvakhajjakaṃ, tassa chaviyādīni chinditvā, yathā ṭhitaṃ, kathaṃ, kammaphalaṃ.
Nikamati, nipatti, nicayo, nicarati, nitaraṇaṃ iccatra – ‘‘do dhassa cā’’ti ettha caggahaṇassa bahulatthattā tena ca ggahaṇena yathāpayogaṃ bahudhā ādeso siyā.
In nikamati, nipatti, nicayo, nicarati, nitaraṇaṃ — here, since the word ca in 27 (§20): Do dhassa ca has the sense of bahula (“variously”), by that word ca there may be replacement in many ways, according to usage.
Yathā – ni upasaggato kamu pada ci cara tarānaṃ paṭhamassa vaggadutiyo iminā dvittaṃ, nikkhamati, nipphatti, nicchayo, niccharati, nittharaṇaṃ.
For example, after the prefix ni the first letter of [the roots] kamu, pada, ci, cara, tara becomes the second letter of its class, and by this rule it is doubled.
lemma
derivation
rule
meaning
▶️🤟⨀(nikkhamati)
ni + kamati → ni + (k→kh)amati → ni + (kh→kkh)amati → nikkhamati
§20, §29
goes out
🚺①⨀(nipphatti)
ni + patti → ni + (p→ph)atti → ni + (ph→pph)atti → nipphatti
§20, §29
accomplishment
🚹①⨀(nicchaya)
ni + cayo → ni + (c→ch)ayo → ni + (ch→cch)ayo → nicchayo
§20, §29
decision
▶️🤟⨀(niccharati)
ni + carati → ni + (c→ch)arati → niccharati
§20, §29
issues forth
🚻①⨀(nittharaṇa)
ni + taraṇaṃ → ni + (t→th)araṇaṃ → nittharaṇaṃ
§20, §29
crossing over
Tathā bo vassa kuva diva siva vajādīnaṃ dvirūpassāti va kāradvayassa ba kāradvayaṃ, yathā – kubbanto, evaṃ kubbāno, kubbanti, sadhātvantayādesassa dvittaṃ. Divati dibbati, evaṃ dibbanto, sibbati, sibbanto, pavajati pabbajati, pabbajanto, nivānaṃ nibbānaṃ, nibbuto, nibbindati, udabbayaṃ iccādi.
Likewise b for v: for the doubled v of [the roots] kuva, diva, siva, vaja and the like — that is, for the pair of vs — a pair of bs; for example kubbanto, and so kubbāno, kubbanti, where the replacement of ya together with the root-final is doubled. divati becomes dibbati, and so dibbanto; sibbati, sibbanto; pavajati becomes pabbajati, pabbajanto; nivānaṃ becomes nibbānaṃ; nibbuto, nibbindati, udabbayaṃ, and so on.
lemma
derivation
rule
meaning
🚹①⨀(kubbanta)
kuv + ya + anto → ku(vv→bb)anto → kubbanto
§20, §28
doing
🚹①⨀(kubbāna)
ku(vv→bb)āno → kubbāno
§20, §28
doing
▶️🤟⨂(kubbati)
ku(vv→bb)anti → kubbanti
§20, §28
they do
▶️🤟⨀(dibbati)
divati → di(vv→bb)ati → dibbati
§20, §28
plays
🚹①⨀(dibbanta)
di(vv→bb)anto → dibbanto
§20, §28
playing
▶️🤟⨀(sibbati)
si(vv→bb)ati → sibbati
§20, §28
sews
🚹①⨀(sibbanta)
si(vv→bb)anto → sibbanto
§20, §28
sewing
▶️🤟⨀(pabbajati)
pavajati → pa(vv→bb)ajati → pabbajati
§20, §28
goes forth
🚹①⨀(pabbajanta)
pa(vv→bb)ajanto → pabbajanto
§20, §28
going forth
🚻①⨀(nibbāna)
nivānaṃ → ni(vv→bb)ānaṃ → nibbānaṃ
§20, §28
nibbāna
🚹①⨀(nibbuta)
ni(vv→bb)uto → nibbuto
§20, §28
quenched
▶️🤟⨀(nibbindati)
ni(vv→bb)indati → nibbindati
§20, §28
becomes disenchanted
🚹②⨀(udabbaya)
uda + vayaṃ → uda(vv→bb)ayaṃ → udabbayaṃ
§20, §28
rise and fall
Lo rassa pari taruṇādīnaṃ kvaci. Paripanno palipanno, evaṃ palibodho, pallaṅkaṃ, taruṇo taluno, mahāsālo, māluto, sukhumālo.
l for the r of pari, taruṇa and the like, in some places: paripanno becomes palipanno, taruṇo taluno, and so the rest.
lemma
derivation
rule
meaning
🚹①⨀(palipanna)
pa(r→l)ipanno → palipanno
§20
sunk into
🚹①⨀(palibodha)
pa(r→l)ibodho → palibodho
§20
an obstruction
🚻①⨀(pallaṅka)
pa(ry→ll)aṅkaṃ → pallaṅkaṃ
§20
the cross-legged posture
🚹①⨀(taluna)
ta(r→l)uno → taluno
§20
tender
🚹①⨀(mahāsāla)
mahāsā(r→l)o → mahāsālo
§20
very wealthy
🚹①⨀(māluta)
mā(r→l)uto → māluto
§20
wind
🚹①⨀(sukhumāla)
sukhumā(r→l)o → sukhumālo
§20
delicate
Ṭo tassa dukkatādīnaṃ kvaci. Yathā – dukkataṃ dukkaṭaṃ, evaṃ sukaṭaṃ, pahaṭo, patthaṭo, uddhaṭo, visaṭo iccādi.
Ko tassa niyatādīnaṃ kvaci. Niyato niyako.
Yo jassa nijādissa vā, nijaṃputtaṃ niyaṃputtaṃ.
Ko gassa kulūpagādīnaṃ, kulūpago kulūpako.
Tathā ṇo nassa pa pariādito. Panidhānaṃ paṇidhānaṃ, evaṃ paṇipāto, paṇāmo, paṇītaṃ, pariṇato, pariṇāmo, ninnayo niṇṇayo, evaṃ uṇṇato, oṇato iccādi.
Likewise ṇ for n after pa, pari and the rest: panidhānaṃ becomes paṇidhānaṃ, ninnayo niṇṇayo, and so the rest.
lemma
derivation
rule
meaning
🚻①⨀(paṇidhāna)
pa(n→ṇ)idhānaṃ → paṇidhānaṃ
§20
aspiration
🚹①⨀(paṇipāta)
pa(n→ṇ)ipāto → paṇipāto
§20
prostration
🚹①⨀(paṇāma)
pa(n→ṇ)āmo → paṇāmo
§20
bowing down
🚻①⨀(paṇīta)
pa(n→ṇ)ītaṃ → paṇītaṃ
§20
excellent
🚹①⨀(pariṇata)
pari(n→ṇ)ato → pariṇato
§20
matured
🚹①⨀(pariṇāma)
pari(n→ṇ)āmo → pariṇāmo
§20
change
🚹①⨀(niṇṇaya)
ni(nn→ṇṇ)ayo → niṇṇayo
§20
decision
🚹①⨀(uṇṇata)
u(nn→ṇṇ)ato → uṇṇato
§20
raised
🚹①⨀(oṇata)
o(n→ṇ)ato → oṇato
§20
bent down
Patiaggi, patihaññati itīdha –
In pati aggi, pati haññati — here,
Kaccāyana puts §48–§50 at the very end of his Sandhikappa, after the niggahīta rules and the go and pā insertions; Buddhappiya gathers them here, with 46 (§79): Tabbiparītūpapade byañjane ca from the Nāmakappa, as the shapes that prefixes and particles take before a consonant — pati to paṭi, putha to puthu, ava to o, o to u — the last topic of his consonant sandhi before the two o rules 47–48. The lead-in from 42 (patiaggi, patihaññati itīdha) shows the words before the change, and his vutti supplies what Kaccāyana’s leaves implicit: that paṭi replaces pati before a vowel as well as before a consonant (sare vā byañjane vā), and that in paṭaggi the i of paṭi then goes by 13 (§12): Sarā sare lopaṃ (pubbassaralopo). kvaci is the exception marker; Ruiz-Falqués counts this rule among the seven in which its role is “not entirely transparent” (Boundaries and Domains §4.1), since the two shapes are not distributed by any phonetic condition but word by word: paṭihaññati yet patilīyati, paṭaggi yet paccattaṃ, where the ti of pati before a vowel has instead gone the way of 22 (§19): Sabbo canti . The counter-example paccattaṃ is Buddhappiya’s own; Kaccāyana’s are paccantimesu, patilīyati, patirūpadesavāso.
Pati iccetassa upasaggassa sare vā byañjane vā pare kvaci paṭi ādeso hoti pubbassaralopo. Paṭaggi, paṭihaññati.
The prefix pati, when a vowel or a consonant follows, is in some places replaced by paṭi; [before a vowel] the preceding vowel is then elided. paṭaggi, paṭihaññati.
lemma
derivation
rule
meaning
🚹①⨀(paṭaggi)
pati + aggi → (pati→paṭi) + aggi → paṭ(i) + aggi → paṭaggi
§48, §12
a counter-fire
🔵▶️🤟⨀(paṭihaññati)
pati + haññati → (pati→paṭi) + haññati → paṭihaññati
§48
is repelled
paṭaggi (4V 5.1: Khuddakavatthu)
Kvacī ti kiṃ? Paccattaṃ, patilīyati.
Puthajano puthabhūtaṃ itīdha ‘‘anto’’ti vattate.
In putha jano, putha bhūtaṃ — here the word anto (the final) continues, from 33 (§43): Pāssa canto rasso .
The second of the prefix shapes: putha before a consonant ends in u. Kaccāyana’s §49 follows §48 just so. Buddhappiya’s gloss adds the doubling of the following j by 40 (§28): Paradvebhāvo ṭhāne (puthujjano) and a counter-example for byañjane, putha ayaṃ, where the a stays before a vowel. The word anto he fetches by a frog’s leap from 33 (§43): Pāssa canto rasso , and its purpose he then argues as Kaccāyana’s vutti had (antaggahaṇena aputhassāpi sare pare antassa ukāro hoti): puthujjano would be obtained even without it, by mere compounding of puthu, which 676 (§666): Puthassa puthupathāmo vā (Kaccāyana §666 (676): Puthassa puthu pathā mo vā ) gives as a form of putha in its own right; so anto must be there to make u the final of a word other than putha too, before a vowel and only in some places — manuññaṃ, evumaṃ, itvevaṃ, the last with its u turned to v by 20 (§18): Vamodudantānaṃ . The sutta has no option marker; the kvaci he applies to this extension is the one continuing from 43 (§48): Kvaci paṭi patissa , and he passes it on again to 45 (‘‘Kvaci byañjane’’ti ca vattate).
Putha iccetassa anto saro u kāro hoti byañjane pare, dvittaṃ. Puthujjano, puthubhūtaṃ.
The final vowel of putha becomes u when a consonant follows; [then] the doubling. puthujjano, puthubhūtaṃ.
lemma
derivation
rule
meaning
🚹①⨀(puthujjana)
putha + jano → puth(a→u) + jano → puthu + (j→jj)ano → puthujjano
§49, §28
an ordinary person
🚻①⨀(puthubhūta)
putha + bhūtaṃ → puth(a→u) + bhūtaṃ → puthubhūtaṃ
§49
widespread
Byañjane ti kiṃ? Putha ayaṃ. ‘‘Puthassa a putha’’iti5
Since [puthujjano] would be established by the compound alone — puthu being a form of putha by ‘‘Puthassa puthu…’’ — the inclusion of anta (the final) by way of a governing word is made again so that, in some places, even the final of a word other than putha becomes u before a vowel.
Mano aññaṃ manuññaṃ, evaṃ imaṃ evumaṃ, paralopo, iti evaṃ itvevaṃ, u kārassa va kāro.
Avakāso, avanaddho, avavadati, avasānamitīdha –
In avakāso, avanaddho, avavadati, avasānaṃ — here,
‘‘Kvaci byañjane’’ti ca vattate.
and kvaci byañjane (in some places, before a consonant) continues.
The third prefix shape, again in Kaccāyana’s own order (§50 after §49). The sutta has no marker of its own; the kvaci byañjane that Kaccāyana’s vutti simply states Buddhappiya obtains by continuation from 44 (§49): Puthassu byañjane and 43 (§48): Kvaci paṭi patissa , and he adds a counter-example for byañjane — avayāgamanaṃ, avekkhati, where before a vowel ava keeps its a with the inserted y of 34 (§35): Yava ma da na ta ra lā cāgamā or takes the e of 16 (§14): Kvacāsavaṇṇaṃ lutte . kvaci is the exception marker, and Ruiz-Falqués again lists this rule among those in which its role is “not entirely transparent” (Boundaries and Domains §4.1): okāso but avasussatu, and osānaṃ beside avasānaṃ, the shape fixed word by word. The lead-in to 46 leaves avassa and the word o in force.
Ava iccetassa upasaggassa o kāro hoti kvaci byañjane pare. Okāso, onaddho, ovadati, osānaṃ.
The prefix ava becomes o in some places when a consonant follows. okāso, onaddho, ovadati, osānaṃ.
lemma
derivation
rule
meaning
🚹①⨀(okāsa)
ava + kāso → (ava→o) + kāso → okāso
§50
opportunity, room
🚹①⨀(onaddha)
ava + naddho → onaddho
§50
covered over
▶️🤟⨀(ovadati)
ava + vadati → ovadati
§50
advises
🚻①⨀(osāna)
ava + sānaṃ → osānaṃ
§50
ending
Kvacī ti kiṃ? Avasānaṃ, avasussatu.
Byañjane ti kiṃ? Avayāgamanaṃ, avekkhati.
Avagate sūriye, avagacchati, avagahetvā itīdha –
In avagate sūriye, avagacchati, avagahetvā — here,
‘‘Avasse’’ti vattate oggahaṇañca.
avassa (of ava) continues, and the word o as well.
Kaccāyana states this rule in the Nāmakappa (§79), among the changes of the vowel o; Buddhappiya fetches it here as the counterpart of 45 (§50): O avassa — tabbiparīto, “the reverse of that o”, is u — so that the three shapes of ava before a consonant, ava, o and u, stand together. His gloss explains every word of the sutta: tabbiparīta is “the reverse of that” (tassa viparīto), upapada “a word pronounced alongside”, the prefix as the first member of a compound, and “the reverse of o” is simply a name for u; he adds the doubling of the following g by 40 (§28): Paradvebhāvo ṭhāne . The ca he reads as katthaci nivattanattha, “for blocking in some places” — the use of ca that Ruiz-Falqués records at Rūp 155, 226 and 440 (Boundaries and Domains §3.5): it shuts off the kvaci option of 45 for these words, so that uggate, uggacchati, uggahetvā have u and neither o nor ava. Kaccāyana’s vutti had read the same ca as avadhāraṇa, emphasis, to the same effect. The sutta has no vā or kvaci of its own.
Ava saddassa upapade tiṭṭhamānassa tassokārassa viparīto ca hoti byañjane pare.
Of the word ava standing as an upapada [the first member of a compound], that o has its reverse as well, when a consonant follows.
Tassa viparīto tabbiparīto, upoccāritaṃ padaṃ upapadaṃ, okāraviparīto ti ukārassetaṃ adhivacanaṃ. Ca saddo katthaci nivattanattho, dvittaṃ. Uggate sūriye, uggacchati, u ggahetvā.
tabbiparīta is “the reverse of that” (tassa viparīto); an upapada is a word pronounced alongside [another]; “the reverse of the letter o” is a name for the letter u. The word ca serves to block [the o] in some places. [Then] the doubling. uggate sūriye, uggacchati, uggahetvā.
lemma
derivation
rule
meaning
🚹⑦⨀(uggata) 🚹⑦⨀(sūriya)
ava + gate → (ava→o→u) + gate → u + (g→gg)ate → uggate sūriye
§50, §79, §28
when the sun has risen
▶️🤟⨀(uggacchati)
ava + gacchati → u + (g→gg)acchati → uggacchati
§50, §79, §28
rises
🔽(uggahetvā)
ava + gahetvā → u + (g→gg)ahetvā → uggahetvā
§50, §79, §28
having learnt
Atippakho tāva, parasataṃ, parasahassaṃ itīdha –
In atippa kho tāva, para sataṃ, para sahassaṃ — here,
‘‘Āgamo’’ti vattate.
āgamo (insertion) continues, from 34 (§35): Yava ma da na ta ra lā cāgamā .
In Kaccāyana this is one of the three insertion rules §35–§37 — consonants, o, niggahīta — and its vutti says only that an o is inserted before a consonant kvaci. Buddhappiya has already dealt with the consonant insertions at 34 (§35): Yava ma da na ta ra lā cāgamā , carries āgamo from there by a frog’s leap, and names the words: the o comes after atippa and para (atippago, parosataṃ, parosahassaṃ); in atippago the ca of 34 supplies the g as well, and where the inserted o meets a vowel that vowel is elided by 13 (§12): Sarā sare lopaṃ . Kaccāyana’s counter-examples (etha passathimaṃ lokaṃ, andhībhūto ayaṃ loko) he drops: for him kvaci says that the insertion is confined to these words — an exception marker in the plain sense of Boundaries and Domains §4.2 — and no counter-example is needed. The rule sits here, after the ava–o–u group, as the first of two o-before-consonant rules; the second, 48 (§183): Etesamo lope , gives the o of manomayaṃ, for which the lead-in manamayaṃ, ayamayaṃ prepares.
Byañjane pare kvaci okārāgamo hotīti atippa para saddehi okārāgamo ‘‘yavamada’’iccādisutte ca saddena atippa to ga kārāgamo ca.
When a consonant follows, in some places the letter o is inserted — that is, the o is inserted after the words atippa and para; and after atippa the letter g is inserted as well, by the word ca in the rule 34 (§35): Yava ma da na ta ra lā cāgamā .
Atippago kho tāva, parosataṃ, parosahassaṃ, ettha ‘‘sarā sare lopa’’nti pubbassaralopo.
Manamayaṃ, ayamayaṃ itīdha ‘‘manogaṇādīna’’nti vattate.
In manamayaṃ, ayamayaṃ — here the word manogaṇādīnaṃ (of the mano-group and the rest) is understood, from 95 (§181): Manogaṇādito smiṃnānamiā .
A rule from the Nāmakappa (§183), where it follows the two mano-group rules §181–§182 and yields the compound form manomayaṃ. Buddhappiya takes it out of his Nāmakaṇḍa — where 95–97 are §181, §182 and §184 — and makes it the last of his consonant sandhi, the o that closes a mano-group stem before a consonant, pairing it with 47 (§36): Kvaci o byañjane . etesaṃ (“of these”) therefore has to be explained from a rule that in his book comes much later, and he says so: ‘‘manogaṇādīna’’nti vattate, the word is understood from 95 (§181): Manogaṇādito smiṃnānamiā (Kaccāyana §181 (95): Manogaṇādito smiṃnānamiā ). His gloss keeps Kaccāyana’s vibhattilope kate and multiplies the examples — manoseṭṭhā, ayopatto, tapodhano, tamonudo, siroruho, tejokasiṇaṃ, rajojallaṃ, ahorattaṃ, rahogato, then āpodhātu, vāyodhātu for the ādi. The Kaccāyana page reads the rule as obligatory; Buddhappiya restricts it: “by a lion’s glance (sīhagati) at the governing vā” — the vā he fetched at 40 (§28): Paradvebhāvo ṭhāne and used at 41 — “it does not occur here”: manamattena, manacchaṭṭhānaṃ, ayakapallaṃ, tamavinodano, manaāyatanaṃ. Idha na bhavati is the phrase of the asanta domain of a vavatthitavibhāsā (Boundaries and Domains §3.2): the o is fixed in manomayaṃ and never appears in manamattena; the vā distributes rather than opens a choice.
Etesaṃ manogaṇādīnamanto ottamāpajjate vibhattilope kate.
The final of these, the words of the mano-group and the rest, becomes o when the inflection has been elided.
Manomayaṃ, ayomayaṃ, evaṃ manoseṭṭhā, ayopatto, tapodhano, tamonudo, siroruho, tejokasiṇaṃ, rajojallaṃ, ahorattaṃ, rahogato.
manomayaṃ, ayomayaṃ, and in the same way the rest. The first member’s inflection has been elided in the compound, and its final a becomes o.
lemma
derivation
rule
meaning
🚻①⨀(manomaya)
mana + mayaṃ → man(a→o) + mayaṃ → manomayaṃ
§183
made of mind
🚻①⨀(ayomaya)
aya + mayaṃ → ay(a→o) + mayaṃ → ayomayaṃ
§183
made of iron
🚹①⨂(manoseṭṭha)
mana + seṭṭhā → manoseṭṭhā
§183
having mind as chief
🚹①⨀(ayopatta)
aya + patto → ayopatto
§183
an iron bowl
🚹①⨀(tapodhana)
tapa + dhano → tapodhano
§183
one whose wealth is austerity
🚹①⨀(tamonuda)
tama + nudo → tamonudo
§183
dispeller of darkness
🚹①⨀(siroruha)
sira + ruho → siroruho
§183
hair of the head
🚻①⨀(tejokasiṇa)
teja + kasiṇaṃ → tejokasiṇaṃ
§183
the fire kasiṇa
🚻①⨀(rajojalla)
raja + jallaṃ → rajojallaṃ
§183
dust and dirt
🚻①⨀(ahoratta)
aha + rattaṃ → ahorattaṃ
§183
day and night
🚹①⨀(rahogata)
raha + gato → rahogato
§183
gone into seclusion
Ādi saddena āpodhātu, vāyodhātu.
Sīhagatiyā vā dhikārato idha na bhavati, manamattena manacchaṭṭhānaṃ, ayakapallaṃ, tamavinodano, manaāyatanaṃ.
Iti byañjanasandhividhānaṃ niṭṭhitaṃ.
Thus ends the treatment of consonant sandhi.
Atha niggahītasandhi vuccate.
Now niggahīta sandhi is stated.
Taṇhaṃ karo, raṇaṃ jaho, saṃ ṭhito, jutiṃ dharo, saṃ mato itīdha ‘‘niggahīta’’nti adhikāro, ‘‘byañjane’’ti vattate.
In taṇhaṃ karo, raṇaṃ jaho, saṃ ṭhito, jutiṃ dharo, saṃ mato — here niggahītaṃ is the governing word, taken from 58 (§30): Aṃ byañjane niggahītaṃ , and byañjane (before a consonant) continues.
Kaccāyana’s niggahīta rules run §30–§34 and §37–§41, with Aṃ byañjane niggahītaṃ first and the changes after it. Buddhappiya reverses the order: he begins with the changes — the class nasal (49), ñ (50–51), m and d (52) — goes on to elision and insertion (53–57), and keeps the “no change” rule §30 for the end (58), just as he put pakati after the vowel changes. niggahītaṃ is thus a governing word taken from a rule still to come, and byañjane continues from 36 (§23): Sarā pakati byañjane . His gloss adds the phonetic reason for the change — the nasal nearest to the following consonant in place of articulation (ṭhānāsannavasena) — and, above all, the analysis of vā: vavatthitavibhāsatthoyaṃ vāsaddo, this vā is a restricted option, with the three domains that Boundaries and Domains §2.3 and §3.2 describe. It is nicca, obligatory, in taṇhaṅkaro, raṇañjaho, saṇṭhito, jutindharo, sammato, the words of the lead-in; anicca, free (vikappena), in taṅkaroti … dātuṅgato, where taṃ karoti is as good; and asanta (idha na bhavati) in na taṃ kammaṃ kataṃ sādhu, saraṇaṃ gacchāmi. Finally he asks why the sutta says vā at all, when a vā would in any case have been carried from another rule, and answers as Kaccāyana’s vutti does (vāggahaṇena niggahitaṃ kho lakārādeso hoti): the express vā signals a further operation, l for the niggahīta of saṃ and of puṃ before an l — paṭisallīno, pulliṅgaṃ.
Vaggabhūte byañjane pare niggahītaṃ kho vaggantaṃ vā pappotīti nimittānusvarānaṃ ṭhānāsannavasena tabbaggapañcamo hoti. Vavatthitavibhāsatthoyaṃ vā saddo. Tena ‘‘taṇhaṅkaro, raṇañjaho, saṇṭhito, jutindharo, sammato’’tiādīsu niccaṃ.
When a class consonant follows, the niggahīta optionally becomes the final of the class: that is, by the nearness in place [of articulation] of the conditioning consonant and the anusvāra, it becomes the fifth letter of that class. This word vā has the sense of a restricted option (vavatthitavibhāsā). Hence in taṇhaṅkaro, raṇañjaho, saṇṭhito, jutindharo, sammato and the like [the change] is obligatory (nicca).
Taṅkaroti taṃ karoti, taṅkhaṇaṃ taṃkhaṇaṃ, saṅgaho saṃgaho, taṅghataṃ taṃ ghataṃ. Dhammañcare dhammaṃ care, tañchannaṃ taṃ channaṃ, tañjātaṃ taṃ jātaṃ, taññāṇaṃ taṃ ñāṇaṃ. Taṇṭhānaṃ taṃ ṭhānaṃ, taṇḍahati taṃ ḍahati6
Idha na bhavati, na taṃ kammaṃ kataṃ sādhu, saraṇaṃ gacchāmi.
Sati copari vā ggahaṇe vijjhantare vā idha vā ggahaṇakaraṇamatthantaraviññāpanatthaṃ, tena niggahītassa saṃ upasaggapumantassa le la kāro. Yathā – paṭisaṃlīno paṭisallīno, evaṃ paṭisallāṇo, sallakkhaṇā, sallekho, sallāpo, puṃliṅgaṃ pulliṅgaṃ.
And since a vā is present further on, or in another rule [from which it could have been carried], the making of a vā here serves to indicate a further meaning: by it the niggahīta of the prefix saṃ and of the final of puṃ becomes l before an l. For example paṭisaṃlīno becomes paṭisallīno, and so paṭisallāṇo, sallakkhaṇā, sallekho, sallāpo; puṃliṅgaṃ becomes pulliṅgaṃ.
lemma
derivation
rule
meaning
🚹①⨀(paṭisallīna)
paṭisaṃ + līno → paṭisa(ṃ→l)līno → paṭisallīno
§31
secluded
🚹①⨀(paṭisallāṇa)
paṭisaṃ + lāṇo → paṭisallāṇo
§31
seclusion
🚺①⨀(sallakkhaṇā)
saṃ + lakkhaṇā → sallakkhaṇā
§31
discernment
🚹①⨀(sallekha)
saṃ + lekho → sallekho
§31
effacement
🚹①⨀(sallāpa)
saṃ + lāpo → sallāpo
§31
conversation
🚻①⨀(pulliṅga)
puṃ + liṅgaṃ → pu(ṃ→l)liṅgaṃ → pulliṅgaṃ
§31
the masculine gender
paṭisallīno (6D 6.1: Brāhmaṇadūtavatthu)
Paccattaṃ eva, taṃ eva, taṃ hi tassa, evaṃ hi vo7
In paccattaṃ eva, taṃ eva, taṃ hi tassa, evaṃ hi vo — here vā is the governing word.
Kaccāyana’s §32 follows §31 in the same way. The rule sits here because the class nasal (49) and the ñ (50–51) are the two shapes a niggahīta takes by assimilation to what follows, and the palatal ñ before e and h has no class of its own to explain it, so that a separate rule is needed. Buddhappiya’s gloss adds that before e the ñ is doubled by 40 (§28): Paradvebhāvo ṭhāne — Kaccāyana’s examples paccattaññeva, taññeva have the doubling but his vutti does not mention it — and reads the vā carried from 49 (§31): Vaggantaṃ vā vagge as a vavatthitavibhāsā whose domain is the two particles eva and hi: before them the change happens; before any other word beginning with e or h it does not (evametaṃ, evaṃ hoti), an asanta domain in the terms of Boundaries and Domains §3.2, where Kaccāyana’s Vāti kasmā? simply lets both forms stand. Only ñaṃ is carried into 51, not ehe.
E kāraha kāre pare niggahītaṃ kho ña kāraṃ pappoti vā, e kāre ñā desassa dvibhāvo.
When the letter e or h follows, the niggahīta optionally becomes the letter ñ; before e the replacing ñ is doubled.
Paccattaññeva paccattaṃ eva, taññeva taṃ eva, tañhi tassa taṃ hi tassa, evañhi vo evaṃ hi vo8
lemma
derivation
rule
meaning
🚻①⨀(paccatta) 🔼(eva)
paccattaṃ eva → paccatta(ṃ→ñ)eva → paccatta(ñ→ññ)eva → paccattaññeva
§32, §28
individually indeed
🚻①⨀(ta) 🔼(eva)
taṃ eva → ta(ṃ→ñ)eva → ta(ñ→ññ)eva → taññeva
§32, §28
that very
🚻①⨀(ta) 🔼(hi) 🚹⑥⨀(ta)
taṃ hi tassa → ta(ṃ→ñ)hi tassa → tañhi tassa
§32
for that, of his
🔼(evaṃ) 🔼(hi) ⚧️③⨂(tumha)
evaṃ hi vo → eva(ṃ→ñ)hi vo → evañhi vo
§32
for thus, by you
paccattaññeva (6D 1.3.1.1: Sassatavāda)
taññeva (1V 1.1.1.5: Vinītavatthu)
tañhi tassa (9M 1.3: Dhammadāyādasutta)
evañhi vo (3V 2.10: Uposathabhedādi)
Vavatthitavibhāsattā vā saddassa evahi nipātato aññattha na hoti, yathā – evametaṃ, evaṃ hoti.
Since the word vā is a restricted option, [the change] does not occur elsewhere than before the particles eva and hi; for example evametaṃ, evaṃ hoti — the asanta domain. Before etaṃ the niggahīta becomes m by 52 (§34): Madā sare ; before hoti it stays.
evametaṃ (1V 1.1.4.2: Vinītavatthu)
evaṃ hoti (no sandhi)
Saṃyogo, saṃyojanaṃ, saṃyato, saṃyācikāya, yaṃ yadeva, ānantarikaṃ yamāhu itīdha ‘‘ña’’miti vattate.
In saṃyogo, saṃyojanaṃ, saṃyato, saṃyācikāya, yaṃ yadeva, ānantarikaṃ yamāhu — here the word ñaṃ continues.
Kaccāyana’s §33 follows §32 just so; ñaṃ is carried on, and sa is saha, “together with”: the niggahīta and the y together become ñ, doubled. Buddhappiya’s gloss again names the vā a vavatthitavibhāsā and fixes its domain — the niggahīta at the end of the word saṃ (saññogo, saññojanaṃ, saññato, saññācikāya) and a niggahīta before the y of a pronoun (yaññadeva, ānantarikaññamāhu); elsewhere it does not occur, etaṃ yojanaṃ, taṃ yānaṃ, saraṇaṃ yanti — an asanta domain (Boundaries and Domains §3.2), where Kaccāyana’s Vāti kasmā? lets saṃyogo stand as a free alternative. He then turns the very wording of the sutta into evidence for something else: it says sa ye where it should say saha ye, and he takes the dropped syllable as a sign that in the suttantas a letter may be elided for ease of pronunciation — hence paṭisaṅkhā yoniso for paṭisaṅkhāya, sayaṃ abhiññā sacchikatvā for abhiññāya, pariyesanā for pariyesanāya, a licence Kaccāyana never states. The lead-in gathers the niggahīta-before-vowel phrases for 52 (§34): Madā sare .
Niggahītaṃ kho ya kāre pare saha ya kārena ña kāraṃ pappoti vā, ñā desassa dvittaṃ.
When the letter y follows, the niggahīta together with the y optionally becomes the letter ñ; the replacing ñ is doubled.
Saññogo saṃyogo, saññojanaṃ saṃyojanaṃ, saññato saṃyato, saññācikāya saṃyācikāya, yaññadeva yaṃ yadeva, ānantarikaññamāhu ānantarikaṃ yamāhu.
lemma
derivation
rule
meaning
🚹①⨀(saññoga)
saṃyogo → sa(ṃy→ñ)ogo → sa(ñ→ññ)ogo → saññogo
§33, §28
conjunction
🚻①⨀(saññojana)
saṃyojanaṃ → sa(ñ→ññ)ojanaṃ → saññojanaṃ
§33, §28
a fetter
🚹①⨀(saññata)
saṃyato → saññato
§33, §28
restrained
🚺③⨀(saññācikā)
saṃyācikāya → saññācikāya
§33, §28
by one’s own begging
🚻①⨀(ya) 🚻①⨀(ya) 🔼(eva)
yaṃ yadeva → ya(ṃy→ñ)adeva → yaññadeva
§33, §28
whatever it is
🚹②⨀(ānantarika) 🚹②⨀(ya) ⏮🤟⨂(āha)
ānantarikaṃ yamāhu → ānantarika(ṃy→ñ)amāhu → ānantarikaññamāhu
§33, §28
which they call “immediate”
saññogo (26Ps 1.2.1: Assādadiṭṭhiniddesa)
saññācikāya (1V 1.2.6: Kuṭikārasikkhāpada)
yaññadeva (3V 6.14: Pāṭaligāmavatthu)
samādhimānantarikaññamāhu (the Ratanasutta’s line, with samādhiṃ joined before it) (18Kh 6: Ratanasutta)
Vā saddassa vavatthitavibhāsattāva saṃ padantato ca sabbanāmaya kāraparato ca niggahītā aññattha na hoti. Yathā – etaṃ yojanaṃ, taṃ yānaṃ, saraṇaṃ yanti.
Ettha ca ‘‘saha ye cā’’ti vattabbe ‘‘sa ye cā’’ti vacanato suttantesu sukhuccāraṇatthamakkharalopopīti daṭṭhabbaṃ, tena paṭisaṅkhāya yoniso paṭisaṅkhā yoniso, sayaṃ abhiññāya sacchikatvā sayaṃ abhiññā sacchikatvā, pariyesanāya pariyesanātiādi sijjhati.
And here, since the sutta says sa ye ca where saha ye ca should have been said, it is to be understood that in the suttantas a letter is also elided for ease of pronunciation. By this paṭisaṅkhā yoniso for paṭisaṅkhāya yoniso, sayaṃ abhiññā sacchikatvā for sayaṃ abhiññāya sacchikatvā, pariyesanā for pariyesanāya and the like are established.
lemma
derivation
meaning
🔽(paṭisaṅkhāya) 🔼(yoniso)
paṭisaṅkhā(ya) yoniso → paṭisaṅkhā yoniso
reflecting wisely
🔼(sayaṃ) 🔽(abhiññāya) 🔽(sacchikatvā)
sayaṃ abhiññā(ya) sacchikatvā → sayaṃ abhiññā sacchikatvā
having realised it by his own direct knowledge
🚺③⨀(pariyesanā)
pariyesanā(ya) → pariyesanā
by seeking
paṭisaṅkhā yoniso (9M 1.2.2: Saṃvarāpahātabbaāsava)
sayaṃ abhiññā sacchikatvā (1V Veranjakanda: Verañjakaṇḍa)
Taṃ ahaṃ brūmi, yaṃ āhu, dhanaṃ eva, kiṃ etaṃ, nindituṃ arahati, yaṃ aniccaṃ, taṃ anattā, etaṃ avoca, etaṃ eva iccatra –
In taṃ ahaṃ brūmi, yaṃ āhu, dhanaṃ eva, kiṃ etaṃ, nindituṃ arahati, yaṃ aniccaṃ, taṃ anattā, etaṃ avoca, etaṃ eva — here,
Kaccāyana’s §34 ends his group of niggahīta changes (§30–§34), and Buddhappiya likewise closes the replacements with it before the elisions of 53–54. The sutta has no vā of its own; the Kaccāyana vutti supplies one, and its page treats it as an open option (tamahaṃ beside akkocchi maṃ). Buddhappiya says where the vā comes from — the governing vā (vāsaddādhikāra) of 49 (§31): Vaggantaṃ vā vagge — and again calls it a vavatthitavibhāsā: the d is confined to the niggahīta of the pronouns ya, ta and eta (yadaniccaṃ, tadanattā, etadavoca), and for the rest there is m or nothing — taṃ ahaṃ, etaṃ eva, akkocchi maṃ avadhi maṃ, the Vā ti kiṃ? answers being the anicca cases of Boundaries and Domains §3.2. His second addition is a yogavibhāga: reading madā apart from sare, he lets m stand for a niggahīta before a consonant too, optionally, and so accounts for the recited buddhama saraṇama gacchāmi. The lead-in gathers the phrases for the elision rule that follows.
Niggahītassa kho sare pare ma kārada kārādesā honti vā. Ettha ca vā saddādhikārassa vavatthitavibhāsattā da kāro yateta saddato parasseva.
When a vowel follows, the niggahīta is optionally replaced by the letters m and d. And here, since the governing word vā is a restricted option, the d is only for [the niggahīta] that follows the words ya, ta and eta.
Tamahaṃ brūmi, yamāhu, dhanameva, kimetaṃ, ninditumarahati, yadaniccaṃ, tadanattā, etadavoca, etadeva.
lemma
derivation
rule
meaning
🚹②⨀(ta) ⚧️①⨀(amha) ▶️👆⨀(brūti)
taṃ ahaṃ → ta(ṃ→m)ahaṃ → tamahaṃ brūmi
§34
him I call
🚻②⨀(ya) ⏮🤟⨂(āha)
yaṃ āhu → ya(ṃ→m)āhu → yamāhu
§34
which they have called
🚻①⨀(dhana) 🔼(eva)
dhanaṃ eva → dhana(ṃ→m)eva → dhanameva
§34
wealth indeed
🚻①⨀(kiṃ) 🚻①⨀(eta)
kiṃ etaṃ → ki(ṃ→m)etaṃ → kimetaṃ
§34
what is this?
🔼(nindituṃ) ▶️🤟⨀(arahati)
nindituṃ arahati → ninditu(ṃ→m)arahati → ninditumarahati
§34
deserves blame
🚻①⨀(ya) 🚻①⨀(anicca)
yaṃ aniccaṃ → ya(ṃ→d)aniccaṃ → yadaniccaṃ
§34
what is impermanent
🚻①⨀(ta) 🚹①⨀(anatta)
taṃ anattā → ta(ṃ→d)anattā → tadanattā
§34
that is not-self
🚻②⨀(eta) ⏮🤟⨀(vacati)
etaṃ avoca → eta(ṃ→d)avoca → etadavoca
§34
he said this
🚻①⨀(eta) 🔼(eva)
etaṃ eva → eta(ṃ→d)eva → etadeva
§34
this very
tamahaṃ brūmi (10M 5.8: Vāseṭṭhasutta)
yamāhu (18It 3.5.9: Dānasutta)
kimetaṃ (6D 3.2.4: Dāsiputtavāda)
ninditumarahati (15A4 1.1.6: Appassutasutta)
yadaniccaṃ (11M 1.6: Āneñjasappāyasutta)
etadavoca (10M 1.1: Kandarakasutta)
Vā ti kiṃ? Taṃ ahaṃ, etaṃ eva, akkocchi maṃ avadhi maṃ.
Ettha ca madā ti yogavibhāgena byañjanepi vā ma kāro. Tena ‘‘buddhama saraṇama gacchāmī’’tiādi sijjhati.
And here, by splitting the rule (yogavibhāga) so that madā stands alone, m [replaces the niggahīta] optionally before a consonant as well. By this buddhama saraṇama gacchāmi and the like are established.
lemma
derivation
rule
meaning
🚹②⨀(buddha) 🚻②⨀(saraṇa) ▶️👆⨀(gacchati)
buddhaṃ saraṇaṃ gacchāmi → buddha(ṃ→m) saraṇa(ṃ→m) gacchāmi → buddhama saraṇama gacchāmi
§34
I go to the Buddha for refuge
Tāsaṃ ahaṃ santike, vidūnaṃ aggaṃ, tassa adāsiṃ ahaṃ itīdha ‘‘sare’’ti vattate.
In tāsaṃ ahaṃ santike, vidūnaṃ aggaṃ, tassa adāsiṃ ahaṃ — here sare (before a vowel) continues.
In Kaccāyana the elision of the niggahīta (§38–§39) comes after the insertion rules §35–§37; Buddhappiya puts it directly after the replacements, keeping sare from 52 (§34): Madā sare , and delays the insertion of a niggahīta to 57 (§37): Niggahitañca , so that his section runs change – elision – insertion – no change. His gloss says where the elision belongs — for the metre and where it eases pronunciation — and, what Kaccāyana’s vutti omits, what happens next: the vowel left before the vanished niggahīta is elided by 13 (§12): Sarā sare lopaṃ , and the following vowel is lengthened by 17 (§15): Dīghaṃ , but only when it does not end in a conjunct (asaṃyogantassa) — tāsāhaṃ, adāsāhaṃ, tathāgatāhaṃ, evāhaṃ, kyāhaṃ, but vidūnaggaṃ. kvaci marks the exception in the plain sense (Boundaries and Domains §4.2): where it does not reach, the niggahīta becomes m by 52, evamassa, kimahaṃ.
Niggahītaṃ kho sare pare lopaṃ pappoti kvaci chandānurakkhaṇe sukhuccāraṇaṭṭhāne. Pubbassaralopo, parassa asaṃyogantassa dīgho.
When a vowel follows, the niggahīta is in some places elided — to keep the metre and where the pronunciation is thereby eased. [Then] the preceding vowel is elided, and the following vowel, when it is not followed by a conjunct, is lengthened.
Tāsāhaṃ santike, vidūnaggaṃ, tassa adāsāhaṃ, tathāgatāhaṃ, evāhaṃ, kyāhaṃ.
lemma
derivation
rule
meaning
🚺⑥⨂(tā) ⚧️①⨀(amha) 🚻⑦⨀(santika)
tāsaṃ ahaṃ → tāsa(ṃ) + ahaṃ → tās(a) + ahaṃ → tās + (a→ā)haṃ → tāsāhaṃ santike
§38, §12, §15
I, in their presence
🚹⑥⨂(vidū) 🚻①⨀(agga)
vidūnaṃ aggaṃ → vidūna(ṃ) + aggaṃ → vidūn(a) + aggaṃ → vidūnaggaṃ (no lengthening before gg)
§38, §12
the foremost of the wise
🚹④⨀(ta) ⏮👆⨀(deti) ⚧️①⨀(amha)
adāsiṃ ahaṃ → adāsi(ṃ) + ahaṃ → adās(i) + ahaṃ → adās + (a→ā)haṃ → tassa adāsāhaṃ
§38, §12, §15
I gave to him
🚹②⨀(tathāgata) ⚧️①⨀(amha)
tathāgataṃ ahaṃ → tathāgata(ṃ) + ahaṃ → tathāgat(a) + (a→ā)haṃ → tathāgatāhaṃ
§38, §12, §15
the Tathāgata, I
🔼(evaṃ) ⚧️①⨀(amha)
evaṃ ahaṃ → ev(a) + (a→ā)haṃ → evāhaṃ
§38, §12, §15
thus I
🚻①⨀(kiṃ) ⚧️①⨀(amha)
kiṃ ahaṃ → ki(ṃ) + ahaṃ → k(i→y)ahaṃ → ky(a→ā)haṃ → kyāhaṃ
§38, §21, §25
what [shall] I
tāsāhaṃ santike (2V 2.2.10: Sikkhaṃpaccācikkhaṇasikkhāpada)
kyāhaṃ (1V 1.2.4.2: Vinītavatthu)
Kvacī ti kiṃ? Evamassa, kimahaṃ.
Ariyasaccānaṃ dassanaṃ, etaṃ buddhānaṃ sāsanaṃ, saṃratto, saṃrāgo, saṃrambho, avisaṃhāro, ciraṃ pavāsiṃ, gantuṃ kāmo, gantuṃ mano itīdha ‘‘kvaci lopa’’nti vattate.
In ariyasaccānaṃ dassanaṃ, etaṃ buddhānaṃ sāsanaṃ, saṃratto, saṃrāgo, saṃrambho, avisaṃhāro, ciraṃ pavāsiṃ, gantuṃ kāmo, gantuṃ mano — here kvaci lopaṃ (in some places, elision) continues.
Kaccāyana’s §39 follows §38 as this follows 53: the ca extends the elision to a following consonant. Buddhappiya’s lead-in from 53 (§38): Kvaci lopaṃ has already put the unelided phrases before the reader (ariyasaccānaṃ dassanaṃ … gantuṃ mano), and his gloss adds the after-effects Kaccāyana leaves unsaid: before r and h the final vowel of the prefix is lengthened by 37 (§25): Dīghaṃ — sāratto, sārāgo, sārambho, avisāhāro — and in cirappavāsiṃ the consonant is doubled by 40 (§28): Paradvebhāvo ṭhāne . kvaci lopaṃ is carried from 53, and kvaci is again the exception marker: etaṃ maṅgalamuttamaṃ keeps its niggahīta before m. Kaccāyana’s second counter-example, taṃ vo vadāmi bhaddante, is dropped.
Niggahītaṃ kho byañjane ca pare lopaṃ pappoti kvaci chandānurakkhaṇādimhi. Ra kāraha kāresu upasaggantassa dīgho.
When a consonant follows too, the niggahīta is in some places elided, for the metre and the like. Before r and h the final of the prefix is lengthened.
Ariyasaccāna dassanaṃ, etaṃ buddhāna sāsanaṃ, sāratto, sārāgo, sārambho, avisāhāro, cirappavāsiṃ, dvittaṃ, gantukāmo, gantumano.
Kvacī ti kiṃ? Etaṃ maṅgalamuttamaṃ.
Kataṃ iti, kiṃ iti, abhinanduṃ iti, uttattaṃ iva, cakkaṃ iva, kaliṃ idāni, kiṃ idāni, tvaṃ asi, idaṃ api, uttariṃ api, dātuṃ api, sadisaṃ eva itīdha ‘‘niggahītamhā, lopa’’nti ca vattate.
In kataṃ iti, kiṃ iti, abhinanduṃ iti, uttattaṃ iva, cakkaṃ iva, kaliṃ idāni, kiṃ idāni, tvaṃ asi, idaṃ api, uttariṃ api, dātuṃ api, sadisaṃ eva — here the words continue as niggahītamhā (after a niggahīta) and lopaṃ (elision).
Kaccāyana’s §40 follows §39 just so: now the niggahīta stays and the vowel after it goes — lopaṃ is carried on, and the inflection form of niggahītaṃ changes to niggahītamhā. Buddhappiya’s gloss adds the consequence: once the vowel has gone the niggahīta stands before a consonant and becomes the class nasal by 49 (§31): Vaggantaṃ vā vagge — katanti, kalindāni, idampi — where Kaccāyana’s abhinandunti, uttattaṃva show it without comment. The vā he once more calls a vavatthitavibhāsā and gives its domain: the particles iti, iva, idāni, asi, api, eva and the like, before which the elision is free — kataṃ iti beside katanti, the Vā ti kiṃ? answers being the anicca cases — while elsewhere it does not occur at all, ahaṃ ettha, etaṃ ahosi, the asanta domain (Boundaries and Domains §3.2). Kaccāyana’s page has only the open option.
Niggahītamhā paro saro lopaṃ pappoti vā. Niggahītassa vaggantattaṃ.
After a niggahīta the following vowel is optionally elided. [Then] the niggahīta becomes the final of the class.
Katanti, kinti, abhinandunti, uttattaṃva, cakkaṃva, kalindāni, kindāni, tvaṃsi, idampi, uttarimpi, dātumpi, sadisaṃva.
Where the consonant that comes to follow the niggahīta is a class letter (t, d, p), the niggahīta becomes the nasal of its class; before v and s it stays.
lemma
derivation
rule
meaning
🚻①⨀(kata) 🔼(iti)
kataṃ iti → kataṃ (i)ti → kata(ṃ→n)ti → katanti
§40, §31
“done”
🚻①⨀(kiṃ) 🔼(iti)
kiṃ iti → kiṃ (i)ti → ki(ṃ→n)ti → kinti
§40, §31
“what?”
⏮🤟⨂(abhinandati) 🔼(iti)
abhinanduṃ iti → abhinandu(ṃ→n)ti → abhinandunti
§40, §31
“they rejoiced”
🚻①⨀(uttatta) 🔼(iva)
uttattaṃ iva → uttattaṃ (i)va → uttattaṃva
§40
like molten [gold]
🚻①⨀(cakka) 🔼(iva)
cakkaṃ iva → cakkaṃ (i)va → cakkaṃva
§40
like a wheel
🚹②⨀(kali) 🔼(idāni)
kaliṃ idāni → kaliṃ (i)dāni → kali(ṃ→n)dāni → kalindāni
§40, §31
now the losing throw
🚻①⨀(kiṃ) 🔼(idāni)
kiṃ idāni → ki(ṃ→n)dāni → kindāni
§40, §31
what now?
⚧️①⨀(tumha) ▶️🤘⨀(atthi)
tvaṃ asi → tvaṃ (a)si → tvaṃsi
§40
you are
🚻①⨀(ima) 🔼(api)
idaṃ api → idaṃ (a)pi → ida(ṃ→m)pi → idampi
§40, §31
this too
🔼(uttariṃ) 🔼(api)
uttariṃ api → uttari(ṃ→m)pi → uttarimpi
§40, §31
further too
🔼(dātuṃ) 🔼(api)
dātuṃ api → dātu(ṃ→m)pi → dātumpi
§40, §31
even to give
🚻①⨀(sadisa) 🔼(eva)
sadisaṃ eva → sadisaṃ (e)va → sadisaṃva
§40
just like
cakkaṃva (18Dh 1.1: Cakkhupālattheravatthu)
tvaṃsi (19Vv 2.3.9: Maṭṭhakuṇḍalīvimānavatthu)
Vā ti kiṃ? Kataṃ iti, kimiti, dātumapi, sāmaṃ eva.
Ayampi vā saddassa vavatthitavibhāsattā itīvīdānisīpevādito aññattha na hoti. Yathā – ahaṃ ettha, etaṃ ahosi.
This too, since the word vā is a restricted option, does not occur elsewhere than before iti, iva, idāni, asi, api, eva and the like — the asanta domain. For example:
lemma
inflected
meaning
⚧️①⨀(amha) 🔼(ettha)
ahaṃ ettha
I, here
🚻①⨀(eta) ⏮🤟⨀(hoti)
etaṃ ahosi
this was
Evaṃ assa te āsavā, pupphaṃ assā uppajjati itīdha sare pare lutte vipariṇāmena ‘‘parasmiṃ, sare, lutte’’ti ca vattate.
In evaṃ assa te āsavā, pupphaṃ assā uppajjati — here, when the following vowel has been elided, the words continue with a change of inflection form as parasmiṃ sare lutte (when the following vowel has been elided).
Kaccāyana’s §41 closes his fourth section as this rule closes the pair on the vowel after a niggahīta; the words niggahītamhā, paro, saro, lutte are carried with changed inflection forms, parasmiṃ sare lutte. Buddhappiya’s gloss defines saññoga, two consonants standing in one place, and then the sense the sutta needs — the first of the two — so that visaññoga is the loss of that first consonant (saṃyogekadesassa purimabyañjanassa lopo): evaṃ assa, evaṃ ssa, evaṃ sa. The ca he reads as Kaccāyana’s vutti does, for a conjunct of three consonants of like form, agyāgāraṃ, vutyassa, where the gg and tt before y lose a member although no niggahīta precedes. The counter-example for lutte, evamassa, is Kaccāyana’s. There is no option marker: the rule is obligatory where its conditions are met. The lead-in gathers the phrases for the insertion of a niggahīta.
Niggahītamhā parasmiṃ sare lutte byañjano saññogo ce, visaññogova hotīti saṃyogekadesassa purimabyañjanassa lopo.
When the vowel following a niggahīta has been elided, if the consonant [that then follows] is a conjunct, it becomes non-conjunct: that is, the former consonant, one member of the conjunct, is elided.
Dvinnaṃ byañjanānamekatra ṭhiti saññogo, idha pana saṃyujjatīti saññogo, purimo vaṇṇo, vigato saññogo assāti visaññogo, paro.
A saññoga is two consonants standing in one place; here, however, saññoga is “that which is conjoined”, the former letter, and visaññoga “that whose conjunct partner has gone”, the latter.
Evaṃ sa te āsavā, pupphaṃsā uppajjati.
lemma
derivation
rule
meaning
🔼(evaṃ) 🚹④⨀(ta) 🚹①⨂(ta) 🚹①⨂(āsava)
evaṃ assa → evaṃ (a)ssa → evaṃ (s)sa → evaṃ sa te āsavā
§40, §41
thus for him those taints
🚻①⨀(puppha) 🚺⑥⨀(ta) ▶️🤟⨀(uppajjati)
pupphaṃ assā → pupphaṃ (a)ssā → pupphaṃ (s)sā → pupphaṃsā uppajjati
§40, §41
her menstruation arises
evaṃsa te āsavā (9M 1.2.2: Saṃvarāpahātabbaāsava)
pupphaṃsā uppajji (the Tipiṭaka’s wording of Buddhappiya’s pupphaṃsā uppajjati) (1V 1.1.1.1: Sudinnabhāṇavāra)
Lutte ti kiṃ? Evamassa.
Ca saddena tiṇṇaṃ byañjanānaṃ sarūpasaṃyogopi visaññogo hoti. Yathā – agyāgāraṃ, vutyassa.
Cakkhu udapādi, ava siro, yāvacidha bhikkhave, aṇuthūlāni, ta sampayuttā itīdha9
In cakkhu udapādi, ava siro, yāva cidha bhikkhave, aṇu thūlāni, ta sampayuttā — here the words sare (before a vowel), āgamo (insertion), kvaci (in some places) and byañjane (before a consonant) continue.
In Kaccāyana this is the third insertion rule (§37), right after Kvaci o byañjane, and its vutti adds by ca a p for vi (pacessati). Buddhappiya moves it to the niggahīta section — the thing inserted is a niggahīta — and drops the vi/pa remark. sare, āgamo, kvaci and byañjane are all carried words, fetched from as far back as 34 (§35): Yava ma da na ta ra lā cāgamā and 36 (§23): Sarā pakati byañjane , and the ca of the sutta adds the niggahīta to the list of insertions. His gloss adds a phonetic limit that Kaccāyana does not state: since the niggahīta is a follower of a short vowel (rassānuga, as the verse at 10 (§8): Aṃiti niggahītaṃ had it), the insertion is only after a short vowel — cakkhuṃ, avaṃ, yāvañ, aṇuṃ, taṃ. kvaci is the exception marker: the insertion is for ease of pronunciation in the places where the texts have it, and na hi etehi, idha ceva are places where they do not. The lead-in (evaṃ vutte, taṃ sādhu …) sets up the general statement that closes the section.
Niggahītañca āgamo hoti sare vā byañjane vā kvaci sukhuccāraṇaṭṭhāne. Niggahītassa rassānugatattā rassatoyevāyaṃ.
And a niggahīta is inserted, before a vowel or a consonant, in some places, where it eases pronunciation. Since the niggahīta is a follower of a short vowel, this [insertion] is only after a short vowel.
Cakkhuṃ udapādi, avaṃsiro, yāvañcidha bhikkhave, aṇuṃthūlāni, taṃsampayuttā, evaṃ taṅkhaṇe, taṃsabhāvo.
lemma
derivation
rule
meaning
🚻①⨀(cakkhu) ⏮🤟⨀(uppajjati)
cakkhu udapādi → cakkhu + (ṃ) udapādi → cakkhuṃ udapādi
§37
vision arose
🚹①⨀(avaṃsira)
ava siro → ava + (ṃ) siro → avaṃsiro
§37
head downwards
🔼(yāva) 🔼(ca) 🔼(idha) 🚹⓪⨂(bhikkhu)
yāva cidha → yāva + (ṃ) cidha → yāva(ṃ→ñ)cidha → yāvañcidha bhikkhave
§37, §31
and in so far as here, bhikkhus
🚻①⨂(aṇuṃthūla)
aṇu thūlāni → aṇu + (ṃ) thūlāni → aṇuṃthūlāni
§37
fine and coarse
🚹①⨂(taṃsampayutta)
ta sampayuttā → ta + (ṃ) sampayuttā → taṃsampayuttā
§37
associated with that
🚹⑦⨀(taṅkhaṇa)
ta khaṇe → ta + (ṃ) khaṇe → ta(ṃ→ṅ)khaṇe → taṅkhaṇe
§37, §31
at that moment
🚹①⨀(taṃsabhāva)
ta sabhāvo → ta + (ṃ) sabhāvo → taṃsabhāvo
§37
of that nature
cakkhuṃ udapādi (3V 1.6: Pañcavaggiyakathā)
Kvacī ti kiṃ? Na hi etehi, idha ceva.
Evaṃ vutte, taṃ sādhu, ekaṃ samayaṃ bhagavā, aggiṃva sandhamaṃ10
In evaṃ vutte, taṃ sādhu, ekaṃ samayaṃ bhagavā, aggiṃva sandhamaṃ — here, where the operations of elision and replacement would apply, he states [the next rule] as their exception for the general case (yebhuyyena).
Kaccāyana opens his niggahīta rules with this one (§30) and lets the changes follow as its exceptions; Buddhappiya closes with it, and the lead-in says why: where the elisions and replacements of 49–57 would apply, this rule is their exception “for the most part” (yebhuyyena), the statement of the residual case, exactly as 36 (§23): Sarā pakati byañjane closed the vowel section. The aṃ of the sutta he glosses as Kaccāyana’s vutti does not: the a is there only for pronunciation (uccāraṇattho), or else it is elided by 67 (§83): Saralopo’mādesappaccayādimhi saralope tu pakati (Kaccāyana §83 (67): Saralopo’ mādesa paccayādimhi saralope tu pakati ) — either way the rule says that a niggahīta before a consonant is simply the niggahīta. Its niggahītaṃ is the word he has been treating as the governing word of the whole section. There is no option marker: this is the general rule (utsarga) of the section, as the Kaccāyana page has it, and the yebhuyyena of the lead-in — the word Ruiz-Falqués notes at Rūp 36 (Boundaries and Domains §2.2) — leaves room for the kvaci and vā rules that precede it. Kaccāyana’s two examples he expands to four.
Niggahītaṃ kho byañjane pare aṃiti hoti.
The niggahīta, when a consonant follows, is aṃ [that is, it remains the niggahīta].
A kāro uccāraṇattho, ‘‘saralopo’’tiādinā pubbassaralopo vā.
The letter a is for pronunciation; or else the preceding vowel [of aṃ] is elided by the rule beginning saralopo, 67 (§83): Saralopo’mādesappaccayādimhi saralope tu pakati .
Evaṃ vutte, taṃ sādhu, ekaṃ samayaṃ bhagavā, aggiṃva sandhamaṃ.
Idha avuttavisesānampi vuttanayātidesatthamatidesamāha.
Here, in order to extend the method already stated to the particular cases not stated, he gives an extension rule (atidesa).
Kaccāyana’s last sandhi rule (§51) is Buddhappiya’s last too, and both read it as an instruction rather than an operation — the vidhyaṅga of the Kaccāyana page: the prefixes and particles not separately treated are to be formed by the methods already stated. Kaccāyana’s vutti then lists some forty such words, vowel sandhi and consonant sandhi in turn; Buddhappiya drops the list and gives instead two hard cases in which the “method already stated” is stretched to its limit. First, the ca of 27 (§20): Do dhassa ca , which he had already read as bahula at 42 (§29): Vagge ghosāghosānaṃ tatiyapaṭhamā , now licenses metathesis (vipariyaya), the exchange of r, h or n with a neighbouring y or v: pariyudāhāsi to payirudāhāsi, ariyassa to ayirassa, bahvābādho to bavhābādho, na abhineyya to anabhineyya. Second, tadamināpi jānātha from taṃ imināpi jānātha: the i goes by 55 (§40): Paro vā saro , an a comes in by taking tatrākāro on its own out of 39 (§27): Lopañca tatrākāro — a yogavibhāga, since no vowel was elided before a consonant there — and the niggahīta becomes d before it by 52 (§34): Madā sare . There is no option marker; the rule is an extension (atidesa), as the lead-in calls it.
Ye idha amhehi visesato na upadiṭṭhā upasagganipātādayo, tesaṃ vuttayogato vuttanayena sarasandhādīsu vuttanayānusārena rūpasiddhi veditabbā.
For those prefixes, particles and the like which we have not separately treated here, the formation of the forms (rūpasiddhi) is to be understood from the rules already stated, by the method stated, in accordance with the method set out under vowel sandhi and the rest.
Tena ‘‘do dhassa cā’’ti sutte ca saddena pariyādīnaṃ rahanādivaṇṇassa vipariyayo yavādīhi, yathā – pariyudāhāsi payirudāhāsi, ariyassa ayirassa, bahvābādho bavhābādho, na abhineyya anabhineyya.
Thus, by the word ca in the rule 27 (§20): Do dhassa ca , in pari and the like the letters r, h, n and so on exchange places (vipariyaya) with y, v and so on: pariyudāhāsi becomes payirudāhāsi, ariyassa ayirassa, bahvābādho bavhābādho, na abhineyya anabhineyya.
lemma
derivation
rule
meaning
⏮🤟⨀(pariyudāharati)
pariyudāhāsi → pa(riy→yir)udāhāsi → payirudāhāsi
§20
he uttered
🚹④⨀(ayira)
ariyassa → a(riy→yir)assa → ayirassa
§20
to the master
🚹①⨀(bavhābādha)
bahvābādho → ba(hv→vh)ābādho → bavhābādho
§20
sickly
🔼(na) ⏯🤟⨀(abhineti)
na abhineyya → (na→an)abhineyya → anabhineyya
§20
he would not lead astray
payirudāhāsi (7D 6.2: Aṭṭhayathābhuccavaṇṇa)
bavhābādho (11M 3.9: Bālapaṇḍitasutta)
anabhineyya (1V 1.2.6: Kuṭikārasikkhāpada)
‘‘Taṃ imināpi jānāthā’’ti ettha ‘‘paro vā saro’’ti sare lutte ‘‘tatrākāro’’ti yogavibhāgena a kāro niggahītassa da kāro, tadamināpi jānātha iccādi.
In taṃ imināpi jānātha, when the following vowel has been elided by 55 (§40): Paro vā saro , an a is inserted by splitting the rule 39 (§27): Lopañca tatrākāro so that tatrākāro stands alone, and the niggahīta becomes d [by 52 (§34): Madā sare]: tadamināpi jānātha, and so on.
taṃ + imināpi jānātha
rule
→
taṃ + (i)mināpi
§40
→
taṃ + (a)mināpi
§27
→
ta(ṃ→d) + amināpi
§34
→
tadamināpi jānātha
tadamināpi jānātha (4V 7.3.1: Saṃghabhedakathā)
Iti niggahītasandhividhānaṃ niṭṭhitaṃ.
Thus ends the treatment of niggahīta sandhi.
Saññāvidhānaṃ sarasandhi sandhi,
The laying down of terms, the joining that is vowel sandhi, niggahīta in the chapter on sandhi —
Iti rūpasiddhiyaṃ sandhikaṇḍo paṭhamo.
Thus ends the Sandhikaṇḍa, the first chapter of the Rūpasiddhi.
