The Complete Pāli Alphabet: Understanding Its 41 Letters and Their Sounds

PĀLI 101 Pronunciation & Phonetics Series · Article 1 of 16
Ven. Dr. Nivitigala Sumitta Thero (Bhante Sumitta) · Dhamma USA

Introduction

Every student of Pāli begins with the sounds of the language. Before we can read a sutta, recite the Three Refuges with care, or follow a grammatical rule, we need to recognize the letters and learn the distinctions they represent. The traditional order of the alphabet itself teaches something, because it follows the way the sounds are made in the mouth.

This article begins Dhamma USA’s sixteen-part series on Pāli pronunciation and phonetics. It introduces the traditional forty-one-letter system, the eight vowels, the five consonant groups, the remaining consonants, niggahīta, and Roman diacritical marks. Later lessons will examine pronunciation and sound changes more fully.

We will distinguish traditional grammar, modern phonetics, historical linguistic analysis, regional recitation, and writing conventions. These perspectives often agree, but not always, so we keep them apart.

PĀLI 101 video lessons (2026): Watch the full playlist on YouTube

Watch: The Complete Pāli Alphabet (PĀLI 101)

Learning objectives

By the end of this lesson, readers should be able to recognize all forty-one traditional alphabetic units, identify the eight vowels and five consonant groups, distinguish Roman digraphs such as kh from doubled consonant sequences such as kkh, understand the function of the main diacritics, and explain the traditional forty-one-letter count alongside Moggallāna’s alternative forty-three-letter analysis.

1. Pāli is a language, not a single script

Pāli is the principal canonical and commentarial language of Theravāda Buddhism. The traditional designation Māgadhī is associated with the language of the Buddha’s teachings, although the exact historical relationship between literary Pāli and the language or languages spoken in ancient Magadha remains a subject of scholarship. Historically, pāḷi could refer to the canonical text or textual line before becoming an established name for the language.

According to Sri Lankan chronicle tradition, the Tipiṭaka was committed to writing in Sri Lanka around the first century BCE, after a long period of oral transmission. Pāli texts have subsequently been written in Sinhala, Burmese, Thai, Khmer, Lao, Devanāgarī, Roman, and other scripts. The same Pāli text can be written in any of these scripts, though each community may pronounce it a little differently.

This series uses Roman transliteration with Unicode diacritics, the usual convention in scholarship and in international teaching.

2. Why forty-one letters?

The influential Kaccāyana grammatical tradition begins with the rules:

Attho akkharasaññāto.
“Meaning is understood through letters.”

Akkharāpādayo ekacattālisaṃ.
“The letters beginning with a are forty-one.”

The spelling above follows the Chaṭṭha Saṅgāyana text. Other editions spell the number slightly differently (ekacattālīsaṃ, ekacattāḷīsaṃ); when citing a particular edition, use its own reading. These opening rules are traditionally followed by the classification of vowels, consonants, groups, and niggahīta.

Category	Pāli term	Count
Vowels	sara	8
Grouped consonants	vagga	25
Other consonants	avagga	7
Niggahīta (nasal sign)	niggahīta	1
Total	akkhara	41

Thus we may speak of eight vowels, thirty-two ordinary consonants, and one additional nasal sign. In the traditional scheme, niggahīta is included among the remaining consonantal signs, producing the familiar expression “eight vowels and thirty-three consonants.”

Why does Moggallāna count forty-three?

Moggallāna’s grammatical tradition gives forty-three vaṇṇa (letters), treating short e and o separately in addition to their long forms. This is another way of analysing vowel length; modern Roman-script Pāli still writes e and o without extra symbols. We use the forty-one-letter Kaccāyana convention as the foundation of this beginner series.

3. One consonant, sometimes two Roman characters

Ten aspirated consonants are written using two Roman characters:

Group	Aspirated consonants
Ka-vagga	kh, gh
Ca-vagga	ch, jh
Ṭa-vagga	ṭh, ḍh
Ta-vagga	th, dh
Pa-vagga	ph, bh

For example, kh represents one traditional consonantal unit, though it is typed as the two Roman characters k and h. This is not the same as an ordinary sequence of an independent k followed by an independent h.

A doubled or geminated consonant is different. In dukkha, kkh represents k followed by aspirated kh; one possible teaching division is duk-kha. Similarly, buddha contains d followed by dh. In the standard spelling of such geminated aspirated stops, the first member is ordinarily unaspirated.

Word	Meaning	Traditional units	Count	Roman characters
khanti	patience	kh-a-n-t-i	5	6
dukkha	suffering	d-u-k-kh-a	5	6
buddha	awakened one	b-u-d-dh-a	5	6
dhamma	teaching; phenomenon	dh-a-m-m-a	5	6
bhikkhu	monk	bh-i-k-kh-u	5	7

The alphabetic unit kh and the Roman text k+h must therefore be distinguished when counting letters in exercises.

4. The eight vowels (sara)

Short (rassa)	Long (dīgha)
a	ā
i	ī
u	ū
—	e
—	o

In the common traditional analysis, a, i, and u are short, while ā, ī, ū, e, and o are long. The macron (¯) marks the long counterparts of a, i, and u. The letters e and o are traditionally placed in the long-vowel category without macrons in standard Roman transliteration.

English comparisons can only be approximate, and speakers’ accents differ. Listen to a competent Pāli reader rather than relying on English spellings alone.

Vowel	Approximate learner cue	Pāli example	Meaning
a	short open central vowel, sometimes compared with u in but	akusala	unwholesome
ā	longer open a, approximately a in father	ānāpāna	in-breathing and out-breathing
i	short vowel, approximately i in bit	iti	thus
ī	longer close front vowel, approximately ee in see	sīla	ethical conduct
u	short rounded vowel, approximately u in put	upāsaka	male lay follower
ū	longer rounded vowel, approximately oo in food	rūpa	form
e	generally a steady vowel, without the English off-glide in they	deva	deity
o	generally a steady vowel, without the English off-glide in go	loka	world

Vowel quantity matters. Compare bhikkhu (“monk,” nominative singular) and bhikkhū (“monks,” nominative plural). The contrast between pana (“but, however”) and pāna (“drink; drinking”) likewise shows why macrons should not be omitted.

Shorter e and o before clusters

Traditional Pāli instruction commonly describes e and o as shortened when followed by a doubled consonant or consonant cluster: compare deva with mettā, and loka with sotthi. This is how the grammarians and teachers describe it; in practice, the length and quality of these vowels vary between recitation traditions. Article 3 examines the matter more closely.

Vowel length and syllable weight are not identical. A syllable containing a short vowel can nevertheless count as heavy (garu) when it is closed by a consonant or ends in niggahīta. In the syllabification met-tā, the first syllable is heavy despite the shortened realization traditionally attributed to e. Article 12 will explain lahu and garu in detail.

5. The twenty-five grouped consonants (vagga)

Group	Traditional place	Voiceless, unaspirated	Voiceless, aspirated	Voiced, unaspirated	Voiced, aspirated	Nasal
Ka-vagga	kaṇṭhaja (guttural)	k	kh	g	gh	ṅ
Ca-vagga	tāluja (palatal)	c	ch	j	jh	ñ
Ṭa-vagga	muddhaja (retroflex/cerebral)	ṭ	ṭh	ḍ	ḍh	ṇ
Ta-vagga	dantaja (dental)	t	th	d	dh	n
Pa-vagga	oṭṭhaja (labial)	p	ph	b	bh	m

Each row has five consonants. The first four distinguish voicing and aspiration; the fifth is the nasal of that articulatory group. For beginners, aspiration may be understood as an accompanying burst of breath. Voicing involves vibration of the vocal folds. Modern phonetics gives more exact descriptions, particularly for voiced aspirates and the palatal series.

Try saying English kin and skin: the initial k in kin often has more aspiration than the k after s in skin. This helps illustrate the contrast between aspirated kh and unaspirated k, although the analogy is not exact for every sound or accent.

Group	Example	Meaning
Ka-vagga	kamma	action
Ca-vagga	citta	mind
Ṭa-vagga	paṇḍita	wise person
Ta-vagga	dhamma	teaching; phenomenon
Pa-vagga	buddha	awakened one

The relationship between the nasal and the other consonants of its group becomes especially clear in words such as saṅgha, pañca, paṇḍita, santa, and sampadā. These show nasals with matching places of articulation; the historical derivation of each form need not be identical.

6. Seven non-grouped consonants (avagga)

Letter	Traditional place	Approximate pronunciation	Example	Meaning
y	tāluja	y in yes	yathā	as; just as
r	muddhaja	tap or trill in many traditions	rāga	passion; attachment
l	dantaja	approximately l	loka	world
v	dantoṭṭhaja	between v and w in some traditions	vedanā	feeling
s	dantaja	usually a voiceless s	sati	mindfulness
h	kaṇṭhaja	approximately h in hand	hadaya	heart
ḷ	muddhaja	retroflex lateral sound	Veḷuvana	Bamboo Grove

The table gives the traditional categories and rough guides to pronunciation; these are not precise phonetic definitions. For example, r and v vary among recitation traditions. Traditional treatments of h in certain combinations also deserve fuller discussion in Article 9. Standard Pāli uses s as its sibilant rather than the separate Sanskrit sibilant letters ś and ṣ.

7. Niggahīta (ṃ)

Niggahīta is the nasal sign traditionally counted as the forty-first alphabetic unit. In the standard Roman convention used here it is written ṃ, with a dot below m. Some other transliteration conventions write ṁ, with a dot above.

A classical grammatical explanation treats niggahīta as dependent on a preceding short vowel, illustrated by aṃ, iṃ, and uṃ. Examples include Buddhaṃ, kiṃ (“what?”), and bhikkhuṃ (“the monk,” accusative singular). This is the grammarians' description; it does not account for every spelling found in the texts.

Niggahīta does not have one sound in every position: its pronunciation depends on what follows it and on the recitation tradition. In some combinations, sandhi turns it into the nasal of the following consonant's group: saṃ + buddha gives sambuddha. This happens in particular combinations only; a written ṃ does not always change.

The regular consonant m belongs to pa-vagga. Niggahīta is classified separately in the traditional alphabet. Article 10 addresses the phonetic and grammatical distinctions fully.

8. Reading Roman diacritics

Mark	Example letters	What it indicates
Macron (¯)	ā ī ū	long vowel
Dot below consonant	ṭ ḍ ṇ ḷ	retroflex articulation in the standard teaching system
Dot above n	ṅ	velar nasal
Tilde above n	ñ	palatal nasal
Dot below m	ṃ	niggahīta
Dot above m	ṁ	alternative spelling of niggahīta

The aspirated consonants ṭh and ḍh inherit the underdot on their initial consonant. These diacritics distinguish real linguistic forms: atthi means “there is,” while aṭṭhi means “bone”; attha can mean “meaning” or “benefit,” while aṭṭha means “eight.”

Use proper Unicode characters rather than omitting the marks when typing notes, handouts, or quotations. Article 15 will address keyboards and transliteration.

9. All forty-one letters at a glance

Category	Letters
Eight vowels	a ā i ī u ū e o
Ka-vagga	k kh g gh ṅ
Ca-vagga	c ch j jh ñ
Ṭa-vagga	ṭ ṭh ḍ ḍh ṇ
Ta-vagga	t th d dh n
Pa-vagga	p ph b bh m
Avagga	y r l v s h ḷ
Niggahīta	ṃ

8 + 25 + 7 + 1 = 41.

10. Common beginner mistakes

Counting kh, th, or ph as two traditional consonants because each uses two Roman characters.
Confusing kh with the longer sequence kkh.
Reading Pāli ph as English f or Pāli dental th as English th in think.
Ignoring vowel length: bhikkhu and bhikkhū are different forms.
Ignoring underdots: atthi and aṭṭhi are not interchangeable.
Reading Pāli c as English k or s rather than an affricate approximately like the beginning of church without the same aspiration.
Assuming ṃ must always sound exactly like m or exactly like ṅ.
Assuming the Roman alphabet is the original script of Pāli.

11. Practice exercises

Exercise 1: Sort the letters. Put each of these in one category: vowel, grouped consonant, non-grouped consonant, or niggahīta.

ā · kh · y · ṃ · ṇ · e · ḷ · bh · s · u · ñ · r · ṭh · o · v · c · h · ī · ṅ · l

Exercise 2: Count traditional alphabetic units and Roman characters.

sati, mettā, jhāna, aṭṭha, saṅgha, dukkha

Exercise 3: Single aspirated consonant or a longer sequence?

kh in khanti; kkh in dukkha; tth in sotthi; ṭṭh in aṭṭha; dh in dhamma.

Exercise 4: Identify the diacritics.

paññā, saṅgha, nibbāna, paṇḍita, Veḷuvana.

Exercise 5: Traditional quantity of e or o. Classify the relevant vowel in deva, mettā, loka, sotthi, and thera. State whether you are applying the traditional classroom rule or describing an observed living pronunciation.

Exercise 6: Reading and recitation. Read the Three Refuges slowly. Underline each long vowel, circle every niggahīta, then identify any heavy syllables that contain a short vowel.

Buddhaṃ saraṇaṃ gacchāmi.
Dhammaṃ saraṇaṃ gacchāmi.
Saṅghaṃ saraṇaṃ gacchāmi.

12. Answer key

Exercise 1: Vowels: ā, e, u, o, ī. Grouped consonants: kh, ṇ, bh, ñ, ṭh, c, ṅ. Non-grouped: y, ḷ, s, r, v, h, l. Niggahīta: ṃ.

Exercise 2:

Word	Traditional units	Roman characters
sati	4	4
mettā	5	5
jhāna	4	5
aṭṭha	4	5
saṅgha	5	6
dukkha	5	6

Exercise 3: kh: one aspirated consonant; kkh: k + kh; tth: t + th; ṭṭh: ṭ + ṭh; dh: one aspirated consonant.

Exercise 4: paññā: ñ (tilde), ā (macron); saṅgha: ṅ (dot above); nibbāna: ā (macron); paṇḍita: ṇ and ḍ (dots below); Veḷuvana: ḷ (dot below).

Exercise 5: deva: long e; mettā: traditionally shortened e before tt; loka: long o; sotthi: traditionally shortened o before tth; thera: long e. These are traditional teaching classifications, not measurements of recordings.

Exercise 6: Each gacchāmi contains long ā. Each line has niggahīta after the first refuge noun (Buddhaṃ, Dhammaṃ, or Saṅghaṃ) and in saraṇaṃ. The final syllables in words ending in ṃ are heavy even though the preceding vowel a is short. Other closed syllables, such as Bud- in Buddhaṃ, also illustrate syllable weight.

Summary

The Kaccāyana tradition teaches forty-one alphabetic units: eight vowels, twenty-five grouped consonants, seven non-grouped consonants, and niggahīta. Moggallāna offers a forty-three-letter analysis by distinguishing short e and o. Roman transliteration expresses some individual Pāli consonants with digraphs such as kh and dh. The careful use of macrons, underdots, and nasal marks is essential to reading the language accurately. These foundations prepare us for the next lesson on places of articulation.

Next article: Where Pāli Sounds Are Produced: Kaṇṭhaja, Tāluja, Muddhaja, Dantaja, and Oṭṭhaja.