Reference Home Page
Light on Pāḷi Pronunciation
Light on Grammar
The Niruttidīpanī by Ledi Sayadaw is an alternative explanation of the aphorisms of a grammar by the Sri Lankan scholar Ven Moggallāna, He founded one of the main schools of Pāḷi grammatical explanation; the first being the school of Ven. Kaccāyana, most probably also a Sri Lankan; and the third being that of Aggavaṁsa, a Burmese. 01 who flourished around the 12th century, the standard explanation being contained in the Moggallānabyākaraṇaṁ itself.
Sayadaw's book appears to have been one of his early works and is well over 500 pages long in the Burmese edition, and it is therefore a major contribution to our understanding of the language, especially in its medieval manifestation. It is generally agreed that the school follows the Sanskrit grammarians and is also based on the more developed language of the Medieval period. The virtue of Ven. Aggavaṁsa's grammar, Saddanīti, on the other hand, lies in the fact that it deals with the Canonical language.02
Here I am only translating the first eight of the explanations, those which deal with the sound-system and pronunciation in Pāḷi. In addition to the text and translation I have also added in tables and explanatory notes wherever I felt it was necessary or helpful in understanding the original text.
There are a number of obscurities in the original text, which I am not sure I have understood completely, so if anyone has any corrections or further explanations I would be happy to receive them.
The work arose out of a request by Ven. Khemaratana, who was giving a workshop on the Pronunciation of Pāḷi, and wanted some authoritative material illustrating the subject.
Namo tassa Bhagavato Arahato Sammāsambuddhassa
Reverence to him, the Gracious One, the Worthy One, the Perfect Sambuddha
The Text of Light on Grammar
The Commencement of the Book
The Sun (who taught) the eighty-four thousand Teachings
Lokamhi yassa jotanti, nantavaṇṇapabhassarā, 
Which shine in the world, the endless resplendent lustre,
Anantavaṇṇaṁ Sambuddhaṁ – vande niruttipāraguṁ.
The endless lustrous Sambuddha – I worship as one who has mastered grammar.
Saddhammañ-cassa Saṅghañ-ca visuddhavaṇṇabhājanaṁ. 
And also the True Teaching and the Community who share in the pure lustre.
Moggallāno mahāñāṇī, niruttāraññakesarī,
Moggallāna, of great knowledge, the lion in the forest of grammar,
Nadibyākaraṇaṁ nādaṁ Sogatāraññabyāpanaṁ. 
Roared the roar of grammar pervading the Fortunate One's forest.
Tassatthaṁ dīpayissāmi, nānārāsiṁ vibhājayaṁ,
I will throw light on the meaning of that (grammar), in various collections and divisions,
Ogāyha saddasatthāni navaṅgaṁ Satthusāsanan-ti. 
After going deeply into the teachers' words (concerning) the nine parts The Teaching of the Buddha was traditionally divided into nine parts: Suttaṁ (Discourses), Geyyaṁ (Mixed Prose and Verse), Veyyākaraṇaṁ (Expositions), Gāthā (Verses), Udānaṁ (Exalted Utterances), Itivuttakaṁ (Thus-Saids), Jātakaṁ (Birth-Stories), Abbhutadhammaṁ (Wonderful Things) and Vedallaṁ (Catecheses).03 of the Teacher's Dispensation.
The Section on (Euphonic) Junction
The Collection of Terms
The Collection of Important Terms
Vaṇṇo, saro, savaṇṇo, dīgho, rasso, byañjano, vaggo, niggahītaṁ.
Sound, vowel, the similar-vowel, long, short, consonant, group, (pure) nasal.
1. A-ādayo titālīsaṁ vaṇṇā.
Beginning with a there are forty-three sounds.
From a to the (pure) nasal, Bindu = niggahīta, written as aṁ below. 04
tecattālīsakkharā vaṇṇā nāma honti:
there are surely forty-three letters (and) sounds:
Atthaṁ vaṇṇenti pakāsentī ti vaṇṇā akkharā ti ca vuccanti,
The meaning is explained, made manifest, therefore vaṇṇa and akkhara are said. The verb vaṇṇeti is indeed a denominative from the word vaṇṇa; but it seems Sayadaw is also making a connection between pakāsati (from pa+kāś) and akkhara (from a+kṣara), which is a rather fanciful etymology. He employs similar devices below, which are quite impossible to translate into English.05
The forms of names and designations are not perishable,
khayavayaṁ na gacchantī ti akkharā,
it is said akkhara do not go to decay and destruction,
nāmagottaṁ na jīratī ti hi vuttaṁ.
it is therefore said that names do not fade. The word akkhara is made up of the prefix a = non and khara = perishable. There is a belief in late Theravāda philosophy that Sayadaw upheld, that three things are eternal: paññatti (designations, another word for nāma, names here), ākāsa (space) and Nibbāna. This is not in keeping with the earliest texts though, where only the latter is eternal. 06
2. Dasādo sarā.
There are ten vowels at the beginning.
Tesu vaṇṇesu ādimhi dasa vaṇṇā sarā nāma honti.
Beginning with these sounds there are ten sounds known as vowels.
Sayam-eva laddhasarūpā hutvā rājanti virocantī ti sarā.
With their own forms by themselves they shine, they blaze forth, therefore sara is said. This appears to be an attempt to derive the word sara from sa meaning self and rāj meaning ruler, perhaps 'able to stand by themselves' (i.e without consonants) is what is intended.07
3. Dve dve savaṇṇā.
The similar-vowels come in groups of two.
Tesu saresu dve dvesarā savaṇṇā nāma honti:
In these vowels there are groups of two vowels that are known as similar-vowels:
a ā a-vaṇṇo, i ī i-vaṇṇo, u ū u-vaṇṇo, et e et-vaṇṇo, ot o ot-vaṇṇo.
a and ā are a-sounds, i and ī are i-sounds, u and ū are ū-sounds, short e and e are et-sounds, The -t at the end of et and ot, are symbols in Pāḷi grammar indicating that the vowel is short.08 short ot and o are o-sounds.
Samāno vaṇṇo suti etesan-ti savaṇṇā sarūpā ti ca vuccanti,
Through these sounds having the same tone, similar-vowels, similar form, is said,
samānaṁ rūpaṁ suti etesan-ti sarūpā.
these have the same form and tone, therefore they have the same form.
4. Pubbo rasso.
The former (vowels) are short (i.e. a, i, u, et, ot).
Dvīsu dvīsu savaṇṇesu yo yo pubbo hoti so so rasso nāma hoti.
In the succession of the groups of two similar-vowels whichever comes first is known as short.
Rassena kālena vattabbā ti rassā,
Because of taking a short time to speak it is therefore short,
rassakālo nāma akkhidalānaṁ ummisananimmisanasamakālo.
a short time is known to be the time it takes to blink.
Tattha, et ot iti dve ekapadasaṁyoge pare kvaci labbhanti:
Herein, (describing) short e and short o, these two are sometimes found in a word before a conjunct (as in):
Eṭṭhi, seṭṭho, oṭṭho, sotthi.
Ekapadasaṁyoge ti kiṁ?
Why is in a word (before a) conjunct said?
Padantarasaṁyoge pare rassā mā hontū ti:
It will not be short before a conjunct at the end of a word (as in):
Maṁ ce tvaṁ nikhaṇaṁ vane (Jā 2.22)
Putto tyāhaṁ Mahārāja (Jā 1.7) Here Sayadaw is saying -e in ce in the first sentence, and -o in putto in the second retain their length, even though they are positioned before conjuncts, because they are at the end of a word, see the next note.09
Kvacī ti kiṁ?
Why is sometimes said?
Ekapadasaṁyoge pi vaggantesu vā ya-ra-la-vesu vā paresu rassā mā hontū ti:
It will not be short in a conjunct in a word where (it stands before) one of the (nasals) at the end of a group, or before ya ra la and va (as in):
Enti, senti, eyya, bhāseyya, meṇḍo, soṇḍo. Modern grammarians do not agree with this. Wilhelm Geiger (A Pāli Grammar, PTS edition, §2.1 Note 1), writes: “The vowels e and o are of medium length; in closed syllables they are short and in open syllables they are long.” K R Norman in his note on this does not disagree. Ven. Buddhadatta (The New Pali Course, II.1) concurs: “It is to be noted that e and o are to be pronounced short before double or conjunct consonants (as in khettam, bhonto, etc.).” Here, the example bhonto which he gives goes against Ledi Sayadaw's explanation. We are unable to determine this from metrical examples as they would both be heavy metrically.10
5. Paro dīgho.
The latter (vowels) are long (i.e. ā, ī, ū, e, o).
Dvīsu dvīsu savaṇṇesu yo yo paro hoti so so dīgho nāma hoti.
In the groups of two similar-vowels those which come at the end are known as long.
Dīghena kālena vattabbā ti dīghā,
Because of taking a long time to speak it is called long,
dīghakālo nāma rassehi diguṇakālo.
a long time is known to be twice as long as the short (vowels).
6. Kādayo byañjanā.
Ka and so on are consonants.
Tesu vaṇṇesu kādayo bindantā, vaṇṇā byañjanā nāma honti.
In the sounds from ka to the (pure) nasal, these are known as consonant sounds.
Atthaṁ byañjayantī ti byañjanā.
They are said to indicate the meaning, (therefore they are) byañjana.
Te pana suddhā addhamattikā,
By themselves they are half a measure,
together with a short (vowel) they are two half measures,
together with a long (vowel) they are three half measures.
7. Pañcapañcakā vaggā.
There are five pentads in the groups.
Tesu byañjanesu kādim-antā
In the consonants beginning with ka until the end
pañcabyañjanapañcakā vaggā nāma honti.
there are five pentad consonants which are known as groups.
Kādi pañcako kavaggo, cādi ca-vaggo,
The pentad beginning with ka is ka-group, beginning with ca is ca-group,
ṭādi ṭa-vaggo, tādi ta-vaggo, pādi pa-vaggo.
beginning with ṭa is ṭa-group, beginning with ta is ta-group, beginning with pa is pa-group.
Sesā avaggā ti siddhaṁ.
The rest (i.e. ya ra la va sa ha ḷa) are ungrouped.
Vaṇṇuddese ekaṭṭhānikānaṁ byañjanānaṁ
In the recitation of the consonant sounds belonging to one position
vagge samūhe niyuttā ti vaggā.
they are ordered into an aggregate of groups, therefore they are groups.
8. Bindu niggahītaṁ.
The dot (mark indicates) the nasal. This refers to the way the (pure) nasal is written in Burmese (and other Indian languages generally), with a dot or bindu.11
Ante bindumatto vaṇṇo niggahītaṁ nāma.
The sound with only a dot at the end is known as the (pure) nasal. Sometimes the bindu is written in place of the proper nasal sound (ṅ, ñ, ṇ, n, m), but when written at the end of a word it is the (pure) nasal.12
Niggayha gayhati uccāriyatī ti niggahītaṁ.
Being pronounced while restraining (the sound to the nose) It means restraining the sound from going out of the mouth so that it only goes out through the nose.13 it is the (pure) nasal.
The Collection of Important Terms is Finished
The Consonant Usage Collection
Ṭhānaṁ karaṇaṁ payatanaṁ.
Position, articulation and effort.
Cha ṭhānāni: kaṇṭhaṭṭhānaṁ, tāluṭṭhānaṁ, muddhaṭṭhānaṁ,
Six positions: the guttural, the palatal, the cerebral,
dantaṭṭhānaṁ, oṭṭhaṭṭhānaṁ, nāsikaṭṭhānaṁ.
the dental, the labial, the nasal. It will be noticed that this is a very logical arrangement of the sounds by the place of articulation.14
Tesu byattaṁ vadantena:
In those who are experienced when they speak (Pāḷi):
yattha akkhan-ti vuccati, taṁ kaṇṭhaṭṭhānaṁ;
where (the word) akkha is said, that is the guttural;
yattha icchan-ti, taṁ tāluṭṭhānaṁ;
where iccha (is said), that is the palatal;
yattha raṭṭhan-ti, taṁ muddhaṭṭhānaṁ;
where raṭṭha (is said), that is the cerebral;
yattha satthan-ti, taṁ dantaṭṭhānaṁ;
where sattha (is said), that is the dental;
yattha pupphan-ti vuccati, taṁ oṭṭhaṭṭhānaṁ;
where puppha (is said), that is the labial;
in the region of the nose is the nasal.
Katthaci pana uraṭṭhānaṁ, siraṭṭhānaṁ, jivhāmūlaṭṭhānan-ti pi āgataṁ:
But some also come at the position of the chest, the head and at the root of the tongue:
tattha siraṭṭhānaṁ nāma muddhaṭṭhānam-eva,
herein, (those) at the position of the head are called cerebrals,
jivhāmūlaṭṭhānaṁ pana sabbavaṇṇānaṁ sādhāraṇan-ti vadanti.
it is said that all the vowels in general are (pronounced) at the root of the tongue.
There are four ways of articulation:
jivhāmūlaṁ, jivhopaggaṁ, jivhaggaṁ, sakaṭṭhānan-ti.
at the root of the tongue, near the tip of the tongue, at the tip of the tongue, and in its own position. I.e. according to the position of its own group, for instance na is pronounced in the position of the dentals, etc.15
Payatanaṁ catubbidhaṁ: saṁvutaṁ, vivaṭaṁ, phuṭṭhaṁ, īsaṁphuṭṭhan-ti.
There are four types of effort: closed, open, with contact, with slight contact.
Tattha karaṇānaṁ sakasakaṭṭhānehi saddhiṁsaṁvaraṇādiko
Herein, for those (four ways of) articulation through their very own position, being with restraint and so on,
visesākāro saṁvutādi nāma.
they are known by their special attribute as closed, etc.
Tattha, kaṇṭhapadesānaṁ aññamaññaṁ saṅghaṭṭanena uppannā:
Herein, in contact with the region of the throat arise:
a-vaṇṇa, ka-vagga, ha-kārā, kaṇṭhajā nāma.
the a-sound, the ka-group (ka kha ga gha ṅa), and the ha-character, known as gutturals.
Tālumhi jivhāmajjhasaṅghaṭṭanena uppannā:
On the palate, with the tongue making contact midway arise:
i-vaṇṇa, ca-vagga, ya-kārā, tālujā nāma.
the i-sound, the ca-group (ca cha ja jha ña) and the ya-character, known as palatals.
Mukhabbhantaramuddhamhi jivhopaggasaṅghaṭṭanena uppannā:
With near the tip of the tongue in contact with the roof of the mouth on the inside arise:
ṭa-vagga, ra-ḷa-kārā muddhajā nāma.
the ṭa-group (ṭa ṭha ḍa ḍha ṇa), ra- and ḷa-characters, known as cerebrals.
Upari dantapantiyaṁ jivhaggasaṅghaṭṭanena uppannā:
With the tip of the tongue in contact with the line of the teeth arise:
ta-vagga, la-sa-kārā dantajā nāma.
the ta-group (ta tha da dha na), and the la- and sa-characters, known as dentals.
In contact with both lips arise:
u-vaṇṇa, pa-vaggā oṭṭhajā nāma.
u-sound, pa-group (pa pha ba bha ma), known as labials.
Niggahītaṁ nāsikajaṁ nāma.
Arising in the nose is the (pure) nasal.
Pañcavaggantā pana nāsikaṭṭhāne pi sakaṭṭhāne pi jāyanti.
The end (sound) of the five groups are nasals produced in their own position.
= (pure) nasal
The e-character is (a diphthong) produced with the throat and the palate.
The o-character is (a diphthong) produced with the throat and lips.
Va-character is produced with the teeth and lips.
Api ca i-vaṇṇ-u-vaṇṇā kaṇṭhe pi jāyanti yeva.
Also i-vowel and u-vowel arise in the throat.
Yadā ha-kāro vaggantehi vā ya, ra, la, vehi vā yutto hoti, tadā urajo ti vadanti.
When ha-character is used together with the groups, or together with ya ra la then they are said to be chest-born (for example):
Pañho, tuṇhi, nhāto, vimhito, gayhate, vulhate, avhānaṁ.
Kaṇṭhaṁ saṁvaritvā uccārito a-kāro saṁvuto nāma.
The a-character is pronounced after having restricted the throat, and is known as closed.
Sakasakaṭṭhāna karaṇāni vivaritvā uccāritā sesasarā ca,
The rest of the vowels are pronounced according to their own positions with open articulation,
sa-ha-kārā ca vivaṭā nāma.
(together with) sa and ha-characters they are known as open (i, u, et, e, ot, o, sa and ha).
Tāni yeva gāḷhaṁ phusāpetvā uccāritā pañcavaggā, phuṭṭhā nāma.
The five groups are pronounced after making close contact, and are known as having contact.
Thokaṁ phusāpetvā uccāritā ya-ra-la-vā, īsaṁphuṭṭhā nāma.
Ya ra la and va are pronounced after making a little contact, and are known as having slight contact.
Tattha oṭṭhajesu tāva pa-vaggaṁ vadantānaṁ
Herein, when with the lips the pa-group (pa pha ba bha ma) is spoken
oṭṭhadvayassa gāḷhaṁ phusanaṁ icchitabbaṁ.
the two lips should have strong contact.
Kasmā? Phuṭṭhapayatanikattā pa-vaggassa.
Why? Pa-group is made with an effort at contact.
U-vaṇṇaṁ vadantānaṁ pana oṭṭhadvayassa vivaraṇaṁ icchitabbaṁ.
But the u-vowel when spoken should have both the lips open.
Kasmā? Vivaṭapayatanikattā u-vaṇṇassa.
Why? The u-vowel is made with an effort at (keeping the lips) open.
Esa nayo sesesu sabbesū ti.
This is the method in all the rest.
Cūḷaniruttiyaṁ pana sabbe rassasarā saṁvutā nāma,
But in the Cūḷanirutti A grammar by Ven. Yamaka in the Kaccāyanapiṭaka.16 it is said all the short vowels are known as closed,
sabbe dīghasarā vivaṭā nāmā ti vuttaṁ,
all the long vowels are known as open,
tathā Saddasāratthajāliniyaṁ katthaci Sakkaṭaganthe ca.
and so in the Saddasāratthajālinī A grammar by Ven. Nāgita.17 and some of the Sanskrit books.
(But) this is the more suitable. It means the Sayadaw feels the description he has given is more suitable than in the other books he mentions.18
Regarding the consonants in the other positions
yuttā sarā attano ṭhānakaraṇāni jahantā pi,
the vowels that are connected are left in their own articulation-positions,
payatanaṁ na jahanti.
but the effort is not given up.
Tasmā nānāvaṇṇānaṁ saṁsagge payatanānaṁ saṁsaggabhedo pi veditabbo ti:
After that in connection with the various sounds the analysis of effort should be understood in this way:
tattha, suṇātu me ti vadanto,
herein, when suṇātu me is said,
yadi ṇā-kāraṁ jivhaggena dantaṭṭhāne katvā vadeyya,
if the -ṇā-character is spoken with the tip of the tongue positioned on the teeth,
dantajo nā-kāro eva bhaveyya.
it will be the -nā-character, a dental.
Tu-kārañ-ca jivhopaggena muddhaṭṭhāne katvā vadeyya,
(If) the -tu-sound (in suṇātu) is spoken near the tip of the tongue on the roof (of the mouth),
muddhajo ṭu-kāro eva bhaveyya.
it will be -ṭu-character, a cerebral.
Evañ-ca sati akkharavipatti nāma siyā.
Thus, in this there will surely be a wrong letter (and sound).
Esa nayo sesesu muddhajadantajesu.
This is the method for the rest of the cerebrals and the dentals.
Tasmā, kammavācaṁ sāventehi,
Therefore, with the declarations of the legal action, These are the legal actions made by Buddhist monastics sitting in session.19
nāma ṭhāna-karaṇa-payatanesu suṭṭhu kusalehi bhavitabban-ti.
the position, articulation and effort should surely be made skilfully. Otherwise the legal action may be considered invalid.20
Sithilañ-ca dhanitañ-ca, dīghaṁ, rassaṁ, garuṁ, lahuṁ,
Unvoiced and voiced, long, short, heavy, light,
niggahītaṁ vimuttañ-ca sambandhañ-ca vavatthitaṁ.
the restrained, the free, the conjunct and the separated.
Mudunā vacīpayogena vattabbā vaggapaṭhama-tatiya-pañcamā sithilā nāma.
The first in the group, the third and the fifth should be spoken with soft application, and are known as unvoiced. Or, unaspirated: ka ga ṅa; ca ja ña, etc.21
Thaddhena vacīpayogena vattabbā vaggadutiya-catutthā dhanitā nāma.
The second and fourth in the group should be spoken with firm application, and are known as voiced. These are what we otherwise call aspirates, as in kha, gha; cha, jha, etc.22
Dīgha-rassā pubbe vuttā.
Long and short are as described previously.
Dīghā ceva saṁyogapubbā ca niggahītantā ca garukā nāma,
Long (vowels), conjunct (consonants) and those with the (pure) nasal at the end are known as heavy,
sesā lahukā nāma.
the rest are known as light. Heavy and light are not the same as long and short, as a short may become heavy if followed by a consonant cluster or a (pure) nasal. This distinction is important for the prosody of the texts.23
Yathā saddasahito vāto mukhachiddena bahi anikkhamma nāsasotābhimukho hoti,
Just as air and sound with an open mouth doesn't go out through the nostrils,
tathā mukhaṁ avivaṭaṁ katvā vattabbaṁ, byañjanaṁ niggahītaṁ nāma.
so you should speak without having opened the mouth, this is known as the (pure) nasal consonant.
Tena yuttāni sabbabyañjanāni niggahītan-tāni nāma,
All the consonants connected with that are known as restrained,
sesā vimuttā nāma.
the rest are known as free.
Padasandhivasena vattabbaṁ sambandhaṁ nāma.
That which is spoken on account of word junction is known as conjunct.
Padacchedaṁ katvā vattabbaṁ vavatthitaṁ nāma.
That which is spoken after splitting the words is known as separated.
The Consonant Usage Collection is Finished
last updated: October 2011