Third Edition Ṭhānuttamo Bhikkhu

A Compendious Grammar on the Language of Pāḷi Buddhism

This eBook is offered freely.
If you wish, you may donate and help us continue offering our services.
May you be happy!

To make a donation, please visit: www.pariyatti.org

PARIYATTI

867 Larmon Road Onalaska, Washington 98570 USA 360.978.4998

www.pariyatti.org

Pariyatti is a nonprofit organization dedicated to enriching the world by:

• v Disseminating the words of the Buddha
• v Providing sustenance for the seeker's journey
• v Illuminating the meditator's path

Māgadhabhāsā (Pāḷi)

A Compendious Grammar on the Language of Pāḷi Buddhism

Māgadhabhāsā (Pāḷi)

A Compendious Grammar on the Language of Pāḷi Buddhism

Ṭhánuttamo Bhikkhu

Third Edition

Pariyatti Press an imprint of Pariyatti Publishing www.pariyatti.org

© Sāsanārakkha Buddhist Sanctuary,

2021

All rights reserved.
No part of this book may be reproduced in any form or by any means without permission in writing from the author.
Permission to reprint this book for free distribution may be granted.

First published, 2021 Second edition,
2022 Revised second edition, 2023 Third edition (Pariyatti),
2024

ISBN 978-1-68172-655-7 (Paperback) ISBN 978-1-68172-698-4 (PDF) Library of Congress Control Number: 2024939806

About Sāsanārakkha Buddhist Sanctuary

Founded in the year 2543 BE (Buddhist Era; i.e. 2000 CE),
it throughout has been the main objective of Sāsanārakkha Buddhist
Sanctuary (SBS) to provide high-quality theoretical and practical tutelage to bhikkhus of the Theravāda.
It is a sanctuary for the protection (ārakkho) of the Buddha's teaching (buddhasāsanaṃ),
heightening knowledge and wisdom in theory and practice.
Suitably for these ends, SBS is situated within an extended primary forest near Taiping city in north-western Malaysia.
Tailored individual spiritual consultations are offered besides
guidance in the form of readings and discussions of the four
main nikāyas and selected texts from the Khuddakanikāya,
periodical vinaya classes and manual skills courses (sewing,
broom making etc.) relating to the craft (sippaṃ) of a bhikkhu.
The rules and regulations as found in the Pāḷi vinaya (monastic
discipline; i.e. the Bhikkhuvibhaṅga and the Khandhakas)
comprise the most fundamental community guidelines for monastics residing at SBS.

Acknowledgments

I would like to thank, first and foremost,
Ā.
Ariyadhammika (Austria) as saṅghanāyaka (
"leader of the community") of Sāsanārakkha Buddhist Sanctuary (SBS),
especially for allowance to pursue my studies in fulltime.
I also wish to express my thanks to Ā.
Bodhirasa (South Africa) for pointing out flaws in the chapters
"Sandhi" and "Morphology" and to Ā.
Pāladhammika (U.S.A.) for reading through a preliminary draft,
giving perceptive feedback.
Sayalay Cālā Therī (Aggācāra International Education and Meditation Centre,
Myanmar) readily responded to numerous of my queries with insightful comments.
I value and recognize her input.
Ā.
Sujāta (Germany) and Ms.
Looi Sow Fei (Malaysia) also went through a draft of the entire book; I am thankful for all the mistakes they spotted.

I appreciate and am grateful for the discussions with Dr. Bryan Levman (University of Toronto,
Canada) about many points and his kind willingness to fully
review an earlier draft version – the quality of this work would have suffered much without his suggestions.
Despite his teaching obligations, Prof. Dr. Thomas Oberlies (Universität Göttingen,
Germany) kindly undertook to review substantial parts of the present grammar; I prize his insightful assistance,
without which it would have suffered a sizable degradation in quality as well.
I wish to thank Dr. Alastair Gornall for his occasional help.
The efforts of Dulip Withanage and his wife Kanchana Ranasinghe (both Sri Lanka) are recognized with gratitude.
They helped with needed book scans from the University of Heidelberg's library,
thus being a prop for the completion of my studies.
Much thanks is also due to Stefan (Germany) and Lamai (Thailand) Köppl who,
in like manner, acted very supportively.
May the spiritual merit (puññaṃ) generated with the creation
and donation of this work be dedicated to the welfare of their recently deceased father (Rudolf Köppl,
dec. 2020) and mother (Malai Namnuan,
dec. 2021) respectively, all beings as a whole and for the longevity of the Buddha's teaching.

Contents

List of Tables

Abbreviations

Ap Apadāna

AN Aṅguttaranikāya

Ay Aniyata

B^e Burmese edition E^e European edition

Bv Buddhavaṃsa

Bv-a Buddhavaṃsa-aṭṭhakathā (Madhuratthavilāsinī)

Cp Cariyāpiṭaka Dhp Dhammapada

Dhp-a Dhammapada-aṭṭhakathā

DN Dīghanikāya

DOP I [Cone, M., 2001]: A dictionary of Pāli (Vol. I) DOP II [Cone,
M., 2013]: A dictionary of Pāli (Vol. II)

It Itivuttaka Jā Jātaka

Jā-a Jātaka-aṭṭhakathā Kacc Kaccāyanabyākaraṇa

Lat.
Latin

Kkh-nṭ Kaṅkhāvitaraṇī-abhinavaṭīkā

MIA Middle Indo-Aryan

Mhv Mahāvaṃsa Mil Milindapañha MN Majjhimanikāya

Mogg Moggallānavyākaraṇaṃ

Moh Mohavicchedanī

Mp Manorathapūraṇī (Aṅguttaranikāya-aṭṭhakathā)

Mp-ṭ Manorathapūraṇīṭīkā (Sāratthamañjūsā)

MW Monier-Williams: A Sanskrit-English dictionary

OIA Old Indo-Aryan Pālim-nṭ Vinayālaṅkāraṭīkā Pār Pārājika

Pd I Paramatthadīpanī I (Udāna-aṭṭhakathā)

Pd III Paramatthadīpanī III (Vimānavatthu-aṭṭhakathā)

Pd IV Paramatthadīpanī IV (Petavatthu-aṭṭhakathā)

Pj I Paramatthajotikā I (Khuddakapāṭha-aṭṭhakathā)

Pj II Paramatthajotikā II (Suttanipāta-aṭṭhakathā)

PED The Pali Text Society's Pali-English dictionary

PTS Pali Text Society Rūp Padarūpasiddhi

Sadd I Saddanītippakaraṇaṃ – padamālā Sadd II Saddanītippakaraṇaṃ – dhātumālā

Skt.
Sanskrit

SN Saṃyuttanikāya Sp Samantapāsādikā

Sv-pṭ Sumaṅgalavilāsinīpurāṇaṭīkā

(Līnatthapakāsinī I; Dīghanikāyaṭīkā)

Th Theragāthā

Vibh-a Vibhaṅga-aṭṭhakathā (Sammohavinodanī)

Vin Vinayapiṭaka

Vism-mhṭ I Visuddhimaggamahāṭīkā (Vol. I)

(Paramatthamañjūsā)

Vmv Vimativinodanīṭīkā

Introduction

Grammar and phonetics are a vital part of the indigenous Buddhist traditions,
right from the era of the Teacher's (i.e. the Buddha's) floruit and throughout history up until modernity,
constituting not only the foundation for preaching the dhamma
to the people but also for understanding the subtleties of it in the first place (Subhūti,
2018: 4).
Thus, we find evidence that those disciplines were invested with
integral significance already in the nearly ubiquitously accepted earliest layers of Buddhist lore,
to quote the Aṅguttaranikāya:

These two things, bhikkhus, lead to the confusion and disappearance of the good dhamma (saddhammo); which two?
Badly- (or "wrongly," "incorrectly") settled words and syllables (or
"letters") and misinterpreted meaning.
Bhikkhus, the meaning of badly-settled words and syllables is misinterpreted [...].
These two things, bhikkhus, lead to the continuance of the good dhamma; which two?
Well-settled words and syllables and well-interpreted meaning.
Bhikkhus, the meaning of well-settled words and syllables is well interpreted (AN II: 7 [AN 2.20]).1

Bearing that in mind, the attempt to elucidate,
elaborate upon and enrich the grammar of the Pāḷi language as
undertaken with the present work seems a meaningful endeavor.

This Māgadhabhāsā (Pāḷi) grammar, as it is named,
was originally not intended to reach the extent it has now.
The initial prospect was to create an informal and more or less
makeshift conglomerate of relevant material mainly for personal studies and general use.
However, the inspiration roused by the thought about the spiritual merit (puññaṃ) gained by creating

¹ Dveme, bhikkhave, dhammā saddhammassa sammosāya antaradhānāya saṃvattanti.
katame dve?
dunnikkhittañca padabyañjanaṃ attho ca dunnīto.
dunnikkhittassa, bhikkhave, padabyañjanassa atthopi dunnayo hoti [...].
dveme, bhikkhave, dhammā saddhammassa ṭhitiyā asammosāya anantaradhānāya saṃvattanti.
katame dve?
sunikkhittañca padabyañjanaṃ attho ca sunīto.
sunikkhittassa, bhikkhave, padabyañjanassa atthopi sunayo hoti [...].

and sharing something more fundamental and reliable by investing
just some extra labor (quite a bit in the end actually) led to
the initial makeshift design being worked upon to lose its rough edges and growing in bulk.

With that, the aims, methods and rationales of the present Pāḷi
grammar are as follows: (a) Lubricating access to the information
contained in numerous modern Pāḷi grammars written in English by collating the dispersed material contained within them.
People who wish to learn about grammatical rules and principles
– either on a broader spectrum or at all – are compelled to track
them down in the thicket of the widely scattered grammar inventories
as separately given by the various available grammars.
These works, mostly fine and outstanding works of scholarship in their own right,
each individually often contain valuable data and perspectives not found in the other ones,
and these are attempted to be distilled and presented with this Pāḷi grammar.
(b) Facilitating identification of and providing explicit reference
to most of the grammatical rules contained in the Kaccāyanabyākaraṇaṃ² (Kaccāyana),
the oldest extant Pāḷi grammar, as well as to selected ones from other traditional grammars.
The complete lack of or just sporadic referencing to the indigenous
grammars – a commendable exception to this being Collins's A
Pali Grammar for Students (2006) – is not a trifling defect.
Not to say that the content which is tendered in such manner is thereby flawed per se,
but it possibly presents disbenefits for a variety of individuals,
such as those who wish to gain familiarity also with the source
grammars or those who are more skeptical by nature about the

² Also Kaccāyanavyākaraṇaṃ: kaccāyana + vyākaraṇaṃ → kaccāyanavyākaraṇaṃ (
"the grammar of Kaccāyana").
The 19th century Sri Lankan scholar bhikkhu Subhūti (2018: 4) explains:
"Vyākaraṇa is the science of writing and speaking a language
without fault and of understanding the intentions of texts by
knowing all the divisions and syntactic relations of a language's expressions.
Vyākaraṇa is not exclusive to one language but is for all languages.
Some languages that were used in former times,
however, do not have complete grammars since the principal users
of these languages were uncivilised and were of weak intellect.
The existence of a very complete and pristine grammatical literature in Sanskrit and Pali,
by contrast, is a testament to the sharp intellects of the users of these languages.
"

validity of unreferenced material.
To my knowledge, neither such a blend as attempted here nor the
consistent referencing to classical grammars has been effected as of this writing,
so that some benefit – however small – might hopefully be derived for the reader from the following pages.
This potential benefit will, it is hoped as well,
not suffer much from the following limitations of the present grammar.

• It does not throughout throw into relief the different ancient
  grammarian's views and presentations (that of Moggallāna,
  Aggavaṃsa etc.).
• Some informative modern grammars have not been taken into consideration.
• It does not deal with prosody.

The structure is primarily modelled after that of Kaccāyana and references (incl.
page numbers) to works in the Pāḷi language as well as quotations
from them are directed to and from the Chaṭṭhasaṅgāyana editions (PDF files) of the Vipassana Research Institute,
Igatpuri, India, also commonly known as the Burmese edition (Be ),
with the exception of two quotations from European editions (E^e ).
Since traditionally proper names and titles of books are not capitalized in the Pāḷi language,
this practice is continued here for the actual Pāḷi texts quoted; however,
it is, for obvious reasons, discontinued for such individual
Pāḷi words contained in the running text written in English.

Those who are not interested in word formation and derivation
but mainly wish to have an avenue quickening access to specific
rules – and thereby to the Pāḷi texts themselves – may skip entire
chapters³ and/or the sections on formation contained within most of the remaining ones.
They may directly proceed to those parts of the book discussing actual usage,
holding the most relevant information for comprehending the
syntax and meaning of the Pāḷi text they wish to understand.
Let it be finally remarked, however, that a proven way to gain a broader and deeper

³ Primarily the chapters
"Sandhi," "Morphology," "Kita and Taddhita Affixes
" and "Uṇādi Affixes."

grasp of the Pāḷi language is to
get also familiar with word formation and derivation principals; therefore,
it is recommended.

Pāḷi – Historical Backdrop

Pāḷi is one of the Middle Indo-Aryan (MIA) languages,
itself part of the Indo-Aryan language family.
The broad classification of Indo-Aryan languages can,
on linguistic grounds,⁴ be chronologically subdivided in the following way(Das,
2006: 3; Gair, 2007: 847; Oberlies,
2007: 164; Witzel, 2009: 47):

• 1. 1500 BCE 600 BCE: Old Indo-Aryan Vedic (Ṛgvedic Sanskrit and its dialects),
     Classical and Epic Sanskrit.
• 1. 600 BC 1200 CE: Middle Indo-Aryan Pāḷi,
     Prākṛt (Prakrit), Ardha-Magadhī, Māharāṣṭrī,
     Gāndhārī, Sinhala Prakrit, Buddhist Hybrid Sanskrit etc.
• 1. 1200 CE present: New Indo-Aryan Hindu/Urdu,
     Sinhala, Dardic, Panjabi, Dogri, Nepali,
     Bengali etc.

The corpora of early Buddhism have initially and in the first
few centuries after the demise of the Teacher been transmitted in four of these Indic languages at a minimum: (1) Pāḷi,
(2) Classical Sanskrit, (3) Gāndhārī and (4) Buddhist Hybrid Sanskrit (Edgerton,
1953: 1).
As the title of this book suggests, only the Pāḷi language will
be singled out and considered in the expositions contained in this grammar.

Basing himself upon morphological and lexical features,
Oberlies (2007: 164) states that Pāḷi cannot be a direct continuation of Vedic,
but Geiger (1916/1956: 1) and Pischel (1957: 4) stress its closer relation to Vedic rather than Classical Sanskrit,
the latter from which Pāḷi, they maintain,
cannot directly be derived.
Wackernagel (as quoted by Karpik, 2019: 55) and others (Karpik:
55; Oberlies: 161) argue for a parallel development of Vedic and the Prakrits in general,
among which Pāḷi and the other Middle Indo-Aryan dialects are sometimes classified (Geiger:

⁴ This classification scheme is not strictly applicable
on historical grounds; MIA languages are older than Classical Sanskrit.

1; Norman,
1983: 7).
Pischel (p.
4) maintains that "[...] it does not seem probable that all the
Prakrit dialects sprang out from one and the same source.
"
Woolner (1999: 3) and von Hinüber (2001: 43),
on the other hand, see them as (essentially) derived from Vedic.
For Levman (2019: 96, n.
32) "the actual answer appears to lie in the middle.
"
In any case, the antiquity of Pāḷi among the Middle Indo-Aryan
languages and largely Sanskritindependent grammatical traditions
were given as reasons to let Pāḷi stand apart from the Prakrits (Klein et al.,
2017: 319).

Pāḷi – Derivation and Orthography

The word "Pāḷi"5 (also "Pāli," "Paḷi
" and anglicized "Pali") was stated to be a Dravidian loanword meaning
"row," "line" and in the Pāḷi Buddhist tradition later also
"norm," "text" (Levman, personal communication,
April 28, 2020; Mallik, 1970: 78, 81; PED,
s.v.
"Pāli/Pāḷi"), being an equivalent to tanti,
meaning
"string," "sacred text."
We also find medieval and modern attempts to derive it from within
the indigenous grammatical system of the Pāḷi language itself as well as from Sanskrit.
The dictionary reference provided above sees a connection with Sanskrit pāli⁶ (
"dam," "dike," "bridge"), but Gombrich (2018: 11) suggests a derivation from Sanskrit √ paṭha (
"to recite") instead.
The autochthonous Pāḷi grammar of Moggallāna,7 however,
maintains a connection to the Pāḷi √pāla,
explained as having the meaning of
"protecting," as it has in Sanskrit8 (Mogg: 147).
It explains: atthaṃ pāti rakkhatīti pāḷi –
"Pāḷi: 'It protects and keeps watch over the meaning.
'
" Childers (s.v.
"Pāli") quotes an anonymous grammatical work being along the
same lines as Moggallāna just referred to: saddatthaṃ pāletīti pāli,
which he renders as:
"Text is so called because it protects the

⁵ IPA: /ˈpaːli/; abbr.
"pi" (ISO 639–1) or "pli" (ISO 639–2 and 3).

⁶ Skt.
√pāla.

⁷ Fl.
ca.
1165 CE (Jayawardhana, 1995: 156).

⁸ Pāto ḷi.
pātismā ḷi hoti.
atthaṃ pāti rakkhatī ti pāḷi tanti –
"ḷi after the [verbal base] pā 'to protect
'.
ḷi occurs after the [verbal base] pā
'to protect'.
It protects (pāti), [i.e.] guards (rakkhati),
the meaning, therefore, it is pāḷi,
[i.e.] a sacred text (tanti)
" (Alastair Gornall [Trans.]; personal communication,
May 4, 2020).

sense of the words.
"
If we accept Jayawickrama's (2003) claim that it is not possible to derive
"Pāḷi" from Sanskrit pāṭha (√paṭha) since no such phonological change is attested,
the constructions of Moggallāna and the dictionary explanations are the preferred choice.

Pāḷi – the Name of a Language

Nowhere in the canon (pāḷi), its commentaries (aṭṭhakathā)
or subcommentaries (ṭīkā) preserved within the Pāḷi tradition is mention made of a language with the name
"Pāḷi."9 In the aṭṭhakathā, it is used solely in the sense of
"text" and predominantly as "canonical text,"
but not exclusively (Norman, 1983: 1; von Hinüber,
1977/1994: 85).
This can be understood from the numerous commentarial expressions
making a clear distinction between canon and aṭṭhakathā.
¹⁰ The aṭṭhakathā and ṭīkā literature termed the
language of the tipiṭaka etc. in the following ways (not exhaustive):11

• Māgadhabhāsā "the language of Magadha
  " (Mp-ṭ II: 178): "Surely, when the the Perfectly Enlightened One committed the Buddha Word,
  the tipiṭaka, to the canon, it was done just by means of the language of Magadha (māgadhabhāsāya).
  "¹²
• Māgadhavohāro "the common (or 'popular,'
  'current) speech of Magadha
  "13 (Kkh-nṭ: 39).

⁹ It is also unknown to non-Buddhist traditions (von Hinüber,
1977/1994: 85).

¹⁰ E.g.: [...] pāḷiyaṃ avuttampi aṭṭhakathāyaṃ vuttavasena gahitaṃ –
"[...] but it is not said in the canon (pāḷi) but taken here vis-à-vis the commentary
" (Sp I: 300).
See also Childers (s.v.
"Pāli") for a decent summary of further occurrances.

¹¹ See also von Hinüber (1977/1994) for further discussion.

¹² Sammāsambuddhopi hi tepiṭakaṃ buddhavacanaṃ tantiṃ āropento māgadhabhāsāya eva āropesi.

¹³ E.g.: ettha ca ariyakaṃ nāma māgadhavohāro.
Levman on the term vohāro (personal communication,
April 28, 2020):
"The word vohāro is derived from OI [Old Indian] vy-ava + hṛ,
meaning 'to carry on business,' 'trade,'
'deal in,' 'exchange,' 'have intercourse with
' etc. In other words,
the very word vohāro confirms the existence of this koine.
"
What this "koine" is referring to is elaborated upon further down below.

• Māgadhiko vohāro "the common speech belonging to Magadha
  "¹⁴ (Sp IV: 23).
• Māgadhikā bhāsā "the language belonging to Magadha
  "15 (Moh: 75).
• Ariyako "Aryan [language]."
• Ariyavohāro "the current Aryan speech
  "16 (Sp I: 94).

This nomenclature landscape makes for the rationale behind selecting the title of the present grammar as it stands,
despite most (but not all) scholars
' dislike of adopting that name for the language in which the
lore of Pāḷi Buddhism was transmitted and in which it has been
committed to writing – a language which was possibly even used
by the Buddha himself (more on that further below in the section
"Pāḷi – What is it?").
How, then, did it come about that we nowadays know that language under the name
"Pāḷi" in the first place and not as it was known throughout,
likely already in the nascent years of Buddhism?

Norman (1983: 1) figures it probable that a misunderstanding of the compound word pāḷibhāsā (
"language of the canon [pāḷi]"), is responsible for the inception of the word
"Pāḷi" as being used to denote the name of a language.
He points to facts indicating a usage of the term
"Pāḷi" in that confused sense in 19th century Sri Lanka and Burma (the work quoted mentioning pāḷibhāsā),
in the case of the latter Burmese work likely even earlier.
Cousins (2015: 119) draws attention to a few ambiguous instances
of pāḷibhāsā in the ṭīkā litarature as denoting the name
of a language but follows eventually von Hinüber (1977/1994: 90),
who made it evident that the first attested use of the word
"Pāḷi," as referring to the language in which the Pāḷi Buddhist scriptures were handed down,
was in the 17th century CE, both pointing to a letter (d.
1672 CE) of M. Laneau (as cited

¹⁴ [S]akāya niruttiyāti ettha sakā nirutti nāma sammāsambuddhena vuttappakāro māgadhiko vohāro.

¹⁵ Sabhāvaniruttīti ca māgadhikā bhāsā,
yāya sammāsambuddhā tepiṭakaṃ buddhavacanaṃ tantiṃ āropenti –
"'The natural tongue': the language belonging to Magadha,
with which the Perfectly Enlightened Ones commit the Buddha Word – the tipiṭaka – to the canon.
"

¹⁶ E.g.: [T]attha ariyakaṃ nāma ariyavohāro,
māgadhabhāsā.

by Pruitt,
1987: 123–4) which relates his successful learning Siam and Pāḷi (Baly).
Lastly, von Hinüber (p.
90) gives the 15th century CE as the terminus ante quem the term
Pāḷi (or pāḷibhāsā) was not used in the sense it is nowadays
and proposes a commencement at some time between the 15th and 17th centuries CE.
He cites a Sinhalese work, which lists four languages among which is also the one of
"Magada" [sic].

For Norman (1983: 2), it seems unlikely that this usage arose independently in all three major Buddhist countries.
In any case, in traditional Buddhist countries both senses –
i.e. the earlier commentarial and the somewhat later confounded
sense – existed parallel to each other up to the present day; thus,
to use the traditional nomenclature as elaborated upon above
as well as the widespread modern variant seem uncontroversial and permissible in nearly all respects.
However, scholars continue to hold certain reservations as to
the legitimacy of allocating much linguistically relevant weight
on a potential link between the Pāḷi language and languages or dialects as they were spoken in ancient Magadha,
at least the Māgadhī dialect proper,
and have proposed quite a plethora of opinions on what Pāḷi
is and how and from what it developed – more on that in the subsequent paragraphs.

Pāḷi – What is it?

The Handbook of Comparative and Historical Indo-European Linguistics (Klein et al.,
2017: 318) states:
"It is generally accepted that Pāli as known from the Theravāda texts was a lingua franca,
not a single individual language particular to one dialect area.
"
However, the scholarly discussions on the subject matter that
have been consulted are of course somewhat more nuanced than
that generalizing statement in its depiction of the status quo.
They state, more specifically, that Pāḷi is either (a) some form
of either a lingua franca,¹⁷ koine18 or standard dialect (Geiger,
1916/1956:

¹⁷ Merriam Webster ("Lingua franca,"
n.d.): "[A]ny of various languages used as common or commercial tongues among peoples of diverse speech.
"

¹⁸ Merriam Webster ("Koine,"
n.d.): "[A] dialect or language of a region that has become the common or standard language of a larger area.
"

4–6; Karpik, 2019: 67; Oberlies; 2007: 183; Roth,
1980: 78; Wynne, 2019: 9–10), (b) some form of a vernacular (Childers,
1875: xiv; Roth, 1980: 78; Warder, 1970/2000: 294) or (c) based upon one of these (Levman,
2019: 64–5, n.
1; Lüders, as quoted by Waldschmidt in Lüders,
1954: 8; Norman, 1989: 66; Rhys Davids,
1911: 53–4).
There is also a dissensus as to the question if Pāḷi predominantly
constitutes an artificially crafted language (Gombrich,
2018: 84–5;19 Norman: 65; von Hinüber,
1996: 520) or had developed mainly by natural means (Pischel,
1957: 5).
It also has to be noted that the first-mentioned views under
(a) above premise some actually spoken basis underlying the Pāḷi language,
having been significantly morphed or superseded by contrived
structures in the course of time – at least in part – and that
the second-mentioned view does not assume that the language was safe from any form of change as it relates to redaction,
transmission errors etc. Not one text-critically involved scholar,
as far as I am aware of, is of the opinion that the Pāḷi as we know it has undergone no changes whatsoever.

The above-presented traditional accounts,
reporting the language as found in the texts of the Pāḷi Buddhist tradition to be māgadhabhāsā etc.,
are by and large considered incorrect by modern scholars.
They adduce, inter alia, the peculiar features of the Māgadhī
dialect proper as inferred from the Aśokan inscriptions and the
medieval descriptions of it by the Indian grammarians and determined
these features to be (a) l instead of r (e.g. lāja – rāja),
(b) a-stems in e for o (e.g. lāje – rājo) and (c) palatal ś for dental s.
However, based upon inscriptional and other evidence,
Norman (1980: 68–9) demonstrated that these features were found
merely within a relatively restricted area and that it is feasible
to regard the home of Pāḷi as being outside the region where the true Māgadhī was spoken but

¹⁹ Gombrich holds that the Buddha was the progenitor
of the Pāḷi language or at least a principle figure as it relates to its creation.

²⁰ Commenting on von Hinüber's assessment of Pāḷi as an artificial language,
Prof. Oberlies remarks:
"The 'artificial language' of Mr. von Hinüber goes too far also for me
" – "Die 'Kunstsprache' von Herrn von Hinüber geht auch mir zu weit
" (personal communication, May 3, 2020).

still within Magadha,
somewhat in the center of the east-Indian region,
not far from Kaliṅga.
He considers it feasible that Māgadhī – as depicted within the
aṭṭhakathā tradition as the language of the tipiṭaka – is
a variant of the Māgadhī dialect proper and that the Buddhist tradition can thus be correct.
To similar conclusions came already Winternitz (1908/1981: 40),
seeing the Māgadhī dialect proper at the base of Pāḷi,
and Geiger (1916/ 1956: 4), to quote the latter:

A consensus of opinion regarding the home of the dialect on which Pāli is based has therefore not been achieved.
Windisch therefore falls back on the old tradition—and I am also
inclined to do the same—according to which Pāli should be regarded as a form of Māgadhī,
the language in which Buddha himself had preached.

What emerges from the above is that the traditional narrative should not be and has not been dismissed outright.

Commentaries, Sub-Commentaries and Pāḷi Grammatical Literature

The aṭṭhakathā and ṭīkā traditions take the language of Magadha
(māgadhabhāsā) to be a natural language – a delightful language indeed (Sv-pṭ: 6).21 As presented already above,
the Samantapāsādikā vinaya aṭṭhakathā (Sp IV: 23) proffers
the following annotation of the phrase sakāya niruttiyā as
used by two Brahmins in the context of one cardinal (as it relates to linguistics) incident recorded in the vinaya,
where they, still attached to things Vedic,
complain about the way or language by adopting or use of which the Buddha's teaching was spoiled:
"[...] 'own tongue' means the common speech belonging to Magadha
(māgadhiko vohāro) in the manner spoken (vuttappakāro) by the Perfectly Enlightened One.
"22 The 12/13th century CE Vimativinodanīṭīkā (Vmv: 125) interprets the relevant portion of the episode thus:
"They ruin (dūsenti) the word of the Buddha with their own
language (sakāya niruttiyā) as it relates to the canon (pāḷi):
'Surely,
those of inferior birth who have learned [memorized; i.e. the buddhavacana]

²¹ Manoramaṃ bhāsanti māgadhabhāsaṃ.

²² [...] sakā nirutti nāma sammāsambuddhena vuttappakāro māgadhiko vohāro.

corrupt the language of Magadha (māgadhabhāsāya) to be spoken by all with ease (sabbesaṃ vattuṃ sukaratāya)
' – this is the meaning.
"²³ The Vinayālaṅkāraṭīkā (Pālim-nṭ: 180) from the 1600
's CE in turn as succinctly as possible glosses sakāya niruttiyā as māgadhabhāsā,
the
"language of Magadha."²⁴ The Samantapāsādikā (Sp I: 94),
on another occasion, equates māgadhabhāsā seemingly with the Aryan language as a whole,
thereby possibly referring to a supra-regional language.25 The
indigenous Pāḷi grammars basically concur with the above.
The Padarūpasiddhi, for example, mentions explicitly that the
Buddha spoke a tongue belonging to Magadha (māgadhika),
as recorded in the tipiṭaka26 (Rūp: 32) – for a detailed discussion
concerning themes related to the last-mentioned point,
see Gornall (2014).
The above is, as we have already seen at the beginning of this chapter,
a sensible account of what language the Buddha employed,
at least primarily.

In this connection, it appears relevant to mention that the aṭṭhakathā
tradition is not just an alternative scholarly opinion but rather constitutes strong additional evidence (cf.
Karpik: 74), as Norman (1983: 119) spelled it out:

[...] some parts of the commentaries are very old,
perhaps even going back to the time of the Buddha,
because they afford parallels with texts which are regarded as canonical by other sects,
and must therefore pre-date the schisms between the sects.
As has already been noted, some canonical texts include commentarial passages,
while the existence of the Old Commentary in the Vinaya-piṭaka
and the canonical status of the Niddesa prove that some sort
of exegesis was felt to be needed at a very early stage of Buddhism.

²³ Pāḷiyaṃ sakāya niruttiyā buddhavacanaṃ dūsentīti
māgadhabhāsāya sabbesaṃ vattuṃ sukaratāya hīnajaccāpi uggaṇhantā dūsentīti attho.

²⁴ Sakāya niruttiyāti māgadhabhāsāya.

²⁵ [T]attha ariyakaṃ nāma ariyavohāro,
māgadhabhāsā.

²⁶ Adhikāroyaṃ.
tattha pañca māre jitavāti jino, buddho.
jinassa vacanaṃ jinavacanaṃ, tassa jinavacanassa yuttaṃ jinavacanayuttaṃ,
tepiṭakassa buddhavacanassa māgadhikāya sabhāvaniruttiyā yuttaṃ anurūpamevāti idaṃ adhikāratthaṃ veditabbaṃ.

Furthermore, Buddhaghosa's Samantapāsādikā contains over 200 quotations of earlier material,
according to the indigenous tradition harkening back in parts
to the first council (paṭhamasaṅgīti) held shortly after the demise of the Buddha (von Hinüber,
1996: 104).
Surely, Geiger (1916/1956: 4–6) must have based his deliberations to some extent upon the exegeses of the aṭṭhakathā,
ṭīkā and grammatical traditions showcased throughout this section when he wrote:

[...] Pāli should be regarded as a form of Māgadhī [...].
Such a lingua franca naturally contained elements of all the dialects [...].
I am unable to endorse the view, which has apparently gained much currency at present,
that the Pāli canon is translated from some other dialect (according to Lüders,
from old Ardha-Māgadhī).
The peculiarities of its language may be fully explained on the
hypothesis of (a) a gradual development and integration of various elements from different parts of India,
(b) a long oral tradition extending over several centuries,
and (c) the fact that the texts were written down in a different country.
I consider it wiser not to hastily reject the tradition altogether
but rather to understand it to mean that Pāli was indeed no pure Māgadhī,
but was yet a form of the popular speech which was based on Māgadhī and which was used by Buddha himself.

Whatever the case may be when it comes to the nature of Pāḷi,
perhaps Bodhi (2020: 3) is right when suggesting:
"If by some unexpected miracle transcripts of the original discourses
should turn up in the exact language(s) in which they were delivered,
one who knows Pāli well would be able to read them with perhaps 90 percent accuracy.
"27 In thus manner, the scope of modern scholarly assessments concerning the nature of Pāḷi partially extends,
but a brief survey of the sociological environment and conditioning
of the Buddha will conclude the account on the nature of Pāḷi as a language with the following section.

²⁷ I am indebted to Ā.
Bodhi (aka Bhikkhu Bodhi) for sharing the above text with me
prior to the book release and Bryan Levman for directing my attention to this passage in the first place.

Pāḷi and the Buddha

The Pāḷi canon does not contain any record about which language the Buddha spoke,
either as his native tongue, regarding potential standard dialects,
a lingua franca or a koine.
As a Sakyan, having possibly been nothing less than
"junior allies"²⁸ of the Kosalan kingdom,
he possibly spoke an eastern Indic dialect as his native tongue
but having received a thoroughgoing education in an aristocratic or royal family,
he in all likelihood was multilingual (cf.
Edgerton, 1953: 2; Karpik, 2019: 21; Levman,
personal communication, April 28, 2020; Rhys Davids,
1911:

This much suffices to understand that
"vassal" is a rendering which misses out on a number of possible nuances.
The respective glosses found in the Sumaṅgalavilāsinī and its ṭīkā make a rendering as
"junior ally" even more compelling.
The former explains anuyuttā with vasavattino (
"wielding power," "dominating"), but the latter clarifies this
term – commenting on the textual variant – to mean anuvattakā (
"siding in with," "one who follows or acts according to
").
Bryan Levman (personal communication,
July 11, 2020) suggest that:
"here vasa must have the meaning of OI vaśa
'willing,
submissive, obedient, subject to or dependent on
' (MW),"
but finds that the traditional exegeses represents a
"commentarial apology" and that it is
"trying to make palatable something unpalatable.
"
It appears to me, however, that the matter,
as pictured above, does not seem to justify probative statements.

²⁸ That this term might be a viable alternative rendering for the commonplace
"vassals" to denote the relationship between the Sakyan crowned
republic and the Kosalan kingdom might be gathered from Pj II (Bodhi [Trans.],
2017: 867):
"[Ruled] by one native to the Kosalans (kosalesu niketino): Saying this,
he rejects its rule by a subordinate ruler; for a subordinate ruler is not said to be native.
But one for whom a particular country has been his place of residence
by way of succession from early times is said to be native to it,
and Suddhodana was such a king.
By this, he shows, too, that it excels in wealth that has come down in succession.
"
DOP II (s.v.
"niketi(n)"): "having a home; (one) who has a long connection with.
"
Although the Dīghanikāya speaks of the Sakyans as being anuyuttā to king Pasenadi of Kosala,
this does not have to refer to vassalage.
The past participle anuyuttā, functioning as a predicate substantive noun in the text,
can mean: (a)
"applying oneself to," "given to," (b)
"following," "attending on," "obedient (as a vassal or inferior king),"
with the textual variant for anuyuttā (i.e. anantarā anuyantā) designating the following: anantarā (
"immediately after," "next"); anuyantā,
from anuyāti (
"to follow," "to go along through,"
"then").

153; Warder, 1970/2000: 200).
There is also evidence that his clan – the Sakyas – spoke Munda
(part of the Austroasiatic language family) and/ or Dravidian (Levman,
2019: 64).
Be that as it may, as Warder (p.
201) and others pointed out, the Buddha spent most of his time
in the kingdom of Kosala and much less so in the Magadhan or others,
and it is outside of Magadha where Buddhism at first in the
main spread,29 although it expanded significantly already during
his lifetime and reached nearly all other ancient Indian countries before the Magadhan supremacy (ca.
410 BCE30 and onwards; Warder: 202).
Thus, although we cannot be certain what kind of language the Buddha habitually employed,
it is at least safe to assume that he was multilingual.
It is also well possible that he made regular use of a more universally
established and widespread form of language – such as a pan-Indic high language,
koine or lingua franca, the existence of which some scholars
have argued for as we have come to see in the foregoing sections.
This indeed might have been the Pāḷi language as preserved in
the voluminous scriptures of Pāḷi Buddhism as we know them today,
described by the tradition throughout under various names,
such as māgadhabhāsā.
Of that language the following grammar is a study.

²⁹ An analysis of the four main nikāyas of the Pāḷi
canon yields the following: 78 % in Kosala and 12 % in Magadha
when including larger cities (such as Sāvatthi); 41 % in Kosala and 20 % in Magadha excluding such (Karpik,
2019: 22–5).

³⁰ Reckoned on the basis of the so-called corrected long chronology,
which places the Buddha's demise at ca.
486 BCE.
This dating remains a viable alternative to the median chronology,
which has gained widespread appeal at present (placing this event at around 400 BCE) and even appears preferable,
especially given the review and symposium of Narain (1993; 2003),
including critical assessments of several highly influential
deductions of scholarly works presented during the Göttingen symposia on the subject (Bechert,
1991; 1992; 1997).

The Pāḷi Alphabet or Orthography (saññā)

There are 41 phonemes to be found in the Pāḷi language,
with the sequential order of them being as follows (Thitzana,
2016: 119): 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 (niggahītaṃ).

(a) The vowel a is appended traditionally to the consonants for ease of utterance,
but a representation without them is also acceptable,
perhaps even preferable (Thitzana, 2016: 121).
(b) By dint of affixing this inherent vowel,
one may also correctly state that the Pāḷi language does not
possess an alphabet – Gornall (2014: 511) called it a syllabary,
but it rather represents a so-called abugida/syllabic alphabet (Ānandajoti,
personal communication, July 15, 2020).

Pāḷi Alphabet Classification

(a) In the traditional classification system,
we find, to facilitate reference, a division into five groups
– called vaggā (pl.) in Pāḷi – of the majority of consonants,
according to the position of the tongue in producing the respective sounds (Ñāṇadhaja,
2011: 9; Thitzana, 2016: 122).
(b) Kaccāyana and other traditional grammars divide all letters into vowels and consonants,
the latter of which includes the niggahītaṃ (Kacc 2–3,
Vidyabhusana, S. & Punnananda, 1935: 3).
(c) One distinct sound is assigned to each letter of the Pāḷi alphabet.
(d) In Roman transliterations of Pāḷi letters,
the aspirates (cha, tha etc.) are represented as digraphs – i.e. two letters standing for one sound (Warder,
1963/2001: 1).
(e) Conjunct consonants (sing.
saṃyogo) are combinations of consonants without intervening vowels,
either with similar (e.g. kkh, dda,
ss) or dissimilar consonants (e.g. ndh,
nd, mba) respectively (Kacc 2–3,
6–7; Thitzana: 53; Vidyabhusana, S. & Punnananda: 3).
(f) Within the traditional Pāḷi grammatical system,
the technical term saṃyogo encompasses also geminate³¹

³¹ Gemination refers to the prolongation in the case of sonorant and fricative consonants (i.a.
/l/, /m/, /n/ and /s/, /z/ respectively) or unreleased consonants in the case of stops (i.a.
/t/, /p/, /d/, /g/; "Gemination," n.d.;"
Geminate consonants," n.d.).

consonants (Rūp: 2–3).
(g) The aspirated letters are not to be regarded as conjunct
consonants since they only represent one separate phoneme or sound each (Ānandajoti,
2004: 16).

Vowels (sarā) – 8

– a, ā, i, ī, u, ū, e, o.

Consonants (byañjanā) – 33

• ka-group (kavaggo) ka, kha,
  ga, gha, ṅa.
• ca-group (cavaggo) ca, cha,
  ja, jha, ña.
• ṭa-group (ṭavaggo) ṭa, ṭha,
  ḍa, ḍha, ṇa.
• ta-group (tavaggo) ta, tha,
  da, dha, na.
• pa-group (pavaggo) pa, pha,
  ba, bha, ma.
• Ungrouped (avaggā) ³² – ya,
  ra, la, va, sa, ha, ḷa.
• aṃ.
• (a) Only the first and second and the third and fourth
  letters of the same class (in that order; e.g. ka + kha but
  not kha + ka) can be conjoined to form a conjunct consonant (here geminates only).
  (b) The fifth letter (nasal) of each class can be appended
  to any consonant of the same classification – including itself – to form conjuncts.
  An exception is the letter ṅ, which cannot form a geminate consonant with itself (Yindee,
  2018: 45).
  (c) The letters of the five groups are articulated with strong contact (cf.
  Rūp: 2; Ñāṇadhaja: 14).

Pāḷi Alphabet – General Descriptions

Vowels

(a) Short (rassaṃ) or light (lahu) are: a,
i, u generally as well as e and o before geminate consonants (kkh,
cch, kk, yy etc.; e.g. bhāseyya –
"He should speak").
Exception for e and o: Occurrences before conjuncts with endgroup nasals are long (e.g. meṇḍo –
"sheep"; soṇḍo – "drunkard"; see above the last letters of each group for the end-group nasals).
(b) According to

³² As per Ñāṇadhaja (2011: 8).

the so-called law of mora, long vowels are usually not followed
by conjunct consonants (one exception out of many is: svākkhāto –
"well taught") – mora being a translation of the Pāḷi term mattā (
"measure").
(c) One mattā denotes the time it takes to pronounce one
short vowel; two mattā it takes for a long one as well as a
short vowel before geminate and conjunct consonants (e.g. nakkhamati –
"He does not approve of," Sp V: 56; Kacc 602; Rūp: 2; Ñāṇadhaja,
2011: 6–8; Oberlies, 2001: 17).
(d) In verse compositions, however,
the short vowels take one and a half mattā to be vocalized:
"In reference (gahaṇa) to light [form]: herein,
however, it might be one and a half mattā as the employed (gahaṇa) use if it relates to verse
"33 (Rūp: 2).
(e) Regarding the long vowels, the same length
"of one and a half mattā is also a well-supported use
"34 (Rūp: 2).
(f) "Short" is said because it takes only a short time to speak one mattā,
corresponding to the time it takes to blink only once (Kacc 4; Rūp: 2; Ñāṇadhaja: 6–8).

(a) Long35 (dīghaṃ) or heavy (garu) are: ā,
ī, ū generally as well as e and o at the end of words (e.g. vane –
"in the forest"; putto – "son"), before single consonants (e.g. kāmesu –
"regarding sensuality"; odanaṃ – "rice
") and, again, the nasal conjuncts mentioned just above (Sp V: 56; Ñāṇadhaja,
2011: 7).
(b) "Long" is said since it takes a long time to articulate two mattā.
(c) The time passage of these long vowels equals that of
two short ones; i.e. to blink twice (Sp V: 6; Ñāṇadhaja: 7).
(d) All vowels are voiced in the Pāḷi language (Duroiselle,
1906/1997: 4).

There are differences in opinion regarding the points just mentioned,
even among the ancient grammarians.
Kaccāyana, for example, mentions e and o as only being long (Kacc 5; Thitzana,
2016: 120), whereas Buddhappiya (Rūp: 12) in his medieval Pāḷi
grammar Padarūpasiddhi – being part of the Kaccāyana tradition
– maintains their shortness before geminate but not conjunct consonants (as above) without allowing exceptions.

³³ Lahuggahaṇañcettha chandasi diyaḍḍhamattassāpi gahaṇatthaṃ.
The short vowel a, i, u are explained to be
"light" (lahu).

³⁴ Aññaggahaṇaṃ diyaḍḍhamattikānampi saṅgahaṇatthaṃ.
The long vowels are referred to as "others
" (añña).

³⁵ The diacritic sign called a
"macron" (¯) above a vowel indicates its length.

However, Moggallāna asserts,
in contradistinction to the Kaccāyana tradition,
that short vowels e and o count as separate letters and
therefore reports a total of 43 phonemes to be existent in the
Pāḷi language.36 See Ñāṇadhaja (2011: 7) for an account of the again differing views held by modern grammarians.
Here we mainly take the explanations of Kaccāyana,
Buddhappiya and Ā.
Ñāṇadhaja (aka Ledi Sayadaw) as the basis for our expositions on phonetics (cf.
Gornall, 2014: 516–7).

With modern examples based upon American English pronunciation (whenever possible),
the following lists tender illustrations of articulating letters
in accordance with the parameters as found in the Pāḷi language.
The letters in parentheses are International Phonetic Alphabet (IPA) symbols (
"Pali," n.d.), modelled after the explanations of the ancient grammarian as to the place (ṭhānaṃ),
instrument (karaṇaṃ) and mode of articulation (payatanaṃ),
given here to broaden the avenues for identification.
The underlined parts of the example words used below correspond phonetically or approximate.

Consonants

(a) Consonants are said to indicate the meaning.
(b) Standing by themselves, they take half a mattā to enunciate,
with a short vowel one and a half mattā and with a long vowel two and a half (Rūp: 2; Ñāṇadhaja,
2011: 8, Oberlies, 2001: 17).
(c) In Pāḷi, there are aspirated (e.g. kha) and unaspirated consonants (e.g. ka).
(d) When letter h arises together with the letters of the groups (represented by an h in the digraphs,
as in kh), or with ya, ra, la,
va, it is said to be chest-born (uraja) – (e) it symbolizes aspiration (Geiger,
1916/1994: 2; Ñāṇadhaja: 12).
Though most English speakers generally do not notice the difference,
there are aspirated and unaspirated sounds in the English language too,
and it might be helpful to briefly elaborate on that at this juncture.
(f) The transcriptions of the

³⁶ "At the end of his discussion,
Moggallāna rejects Buddhappiya's standpoint by arguing that
'e' and 'o' cannot have a long duration and then,
in certain situations, be short.
If 'e' and 'o' are ever short these sounds must be acknowledged as completely different sounds from long
'e' and 'o'
" (Gornall, 2014: 519).

various English words are oftentimes
"broad," which means that they do not disclose as much detail as a so-called
"narrow transcription" would.
With the latter, it becomes possible to discern aspiration in the written word,
as can be seen a little further below.
(g) To give an illustration of what is meant by aspiration in English: When respectively speaking
"skirt" and "keen" closely into the palm of one's hand,
an air puff becomes palpable when producing the [kʰ] in
"keen" (represented by the aspiration modifier letter
"◌ʰ") but not so with the [k] in "skirt
" (showing that it is unaspirated).
To clearly instance the pronunciation of the Pāḷi language,
the relevant English examples are given according to a narrow transcription.
(h) Note that only voiceless stops (p,
t, k) occur as aspirated in English – and this just under certain conditions – not so voiced ones (b,
d, g; Andersen, n.d.).
(i) Double consonants are pronounced as two distinct syllables,
with a noticeable pause (e.g. panna –
"fallen," "gone" ~ "unnecessary"; phassa –
"contact" ~ "pus, sir"; Dhammajoti, 2018: 2; Warder,
1963/2001: 4).

Pāḷi Alphabet: Articulation

Gutturals (kaṇṭhaja)³⁷

(a) The letters of this group are a,
ā, ka, kha, ga, gha, ṅa,
ha and their articulation takes place in the region of the throat,
being gutturals (Rūp: 1; Ñāṇadhaja,
2011: 11) – the end-group nasal is,
in addition to the throat, also produced in the nose (nāsikaṭṭhānaja; Rūp: 2).
(b) Unvoiced (aghosa) 38 letters of the gutturals are: ka,
kha, with the voiced (ghosa) ³⁹ ones being: a,
ā, ga, gha, ṅa, ha (Kacc 9; Duroiselle,
1906/1997: 4; Vidyabhusana, S. & Punnananda,
1935: 5).
(c) As already explained above, the letters of the five groups are articulated with strong contact (cf.
Rūp: 2; Ñāṇadhaja: 14).

^– a [ɐ] = nut.

^– ā [aː] = calm.

^– ka [k] = luck.

³⁷ Lit.
"throat-born."

³⁸ Unvoiced: without vibration of the vocal cords.

³⁹ Voiced: with vibration of the vocal cords.

• – kha [kʰ] ~ keel,
  with stronger breath pulse.
• ga [ɡ] = gear.
• gha [ɡʰ] ~ gear, with breath pulse as with kha.
• ṅa [ŋ] = thing.
• ha [h] = behind.

Palatals (tāluja)⁴⁰

• (a) The letters of this group are i,
  ī, ca, cha, ja, jha, ña,
  ya and their articulation takes place on the palate with the
  tongue's middle (instead of its tip) in contact with it (Rūp: 1–2; Ñāṇadhaja,
  2011: 11) – the end-group nasal is,
  in addition to the palate, also produced in the nose (nāsikaṭṭhānaja; Rūp: 2).
  (b) The letters ca and cha of the palatals are unvoiced but voiced are: ja,
  jha, ña, ya (Kacc 9; Vidyabhusana,
  S. & Punnananda, 1935: 5).
  (c) Letter ya is formed by somewhat touching,
  with slight contact (Rūp: 2; Ñāṇadhaja: 14).
• i [ɪ] = sit.
• ī [iː] = seek.
• ca [tʃ] = which.
• cha [tʃʰ] ~ check, with stronger breath pulse.
• ja [dʒ] = range.
• jha [dʒʱ] ~ range, with breath pulse as with cha.
• ña [ɲ] = señor.
• ya [j] = yes.

⁴⁰ Lit.
"palate-born."

Cerebrals/Retroflexes (muddhaja)⁴¹

(a) The letters of this group are ṭa,
ṭha, ḍa, ḍha, ṇa, ḷa, ra and engendered with near the tip of the tongue,
curled back at the roof of the mouth's interior (Rūp: 2; Ñāṇadhaja,
2011: 11) – the end-group nasal is,
in addition

to that, also produced in the nose (nāsikaṭṭhānaja; Rūp: 2).
(b) Of these, ṭa and ṭha are unvoiced letters,
whereas ḍa, ḍha, ṇa, ḷa, ra are all voiced (Kacc 9).
(c) The letter ra is formed by somewhat touching,
with slight contact (Rūp: 2; Ñāṇadhaja: 14).
(d) The intervocalic letter ḷa usually represents ḍa,
and ḷh may occur for ḍh, but only rarely.
Of ten interchanged are ḷ and l (Collins,
2006: 2).
(e) The way of pronouncing the letters of this class represents the

Figure 1.
Tongue position while pronouncing cerebral/retroflex sounds.
Source: Retroflex consonant (n.d.).
In: Wikipedia – The free encyclopedia.

typical Indian way of articulating the English alveolars,
regular /t/, /d/ etc. (Warder, 1963/2001: 1).
(f) However, when /r/ precedes /t/ and /d/ in American English,
they manifest usually as retroflexes (see below for examples and,
to further aid correct implementation,
Figure 1 above).

• ṭa [ʈ] = heart.
• ṭha [ʈʰ] ~ barter, with stronger breath pulse.
• ḍa [ɖ] = warder.
• ḍha [ɖʰ] ~ warder, with breath pulse as with ṭha.
• ṇa [ɳ] = barn.
• ḷa [ɭ] = curl.
• ra [ɻ] = ram.

Dentals (dantaja)⁴²

(a) The letters of this group are ta,
tha, da, dha, na, la, sa and sounded with the tip
of the tongue in contact with the edge of the row or line of the

⁴¹ Lit.
"head-born."

⁴² Lit.
"tooth-born."

teeth (Rūp: 2; Ñāṇadhaja,
2011: 11) – the end-group nasal is,
in addition to the teeth, also produced in the nose (nāsikaṭṭhānaja; Rūp: 2).
(b) The unvoiced letters here are ta,
tha and sa.
Voiced forms are given as da, dha,
na, la (Kacc 9).
(c) Letter la is formed by somewhat touching,
with slight contact (Rūp: 2; Ñāṇadhaja: 14).
(d) The sibilant sa produces a hissing sound and is unvoiced (Kacc 9; Duroiselle,
1906/1997: 5).
(e) Pāḷi has no [z] sound as found in English (e.g.
"zone"; Warder, 1963/2001: 3).
(f) An English equivalent of tongue positioning for these
dentals would be in the formation of dental fricative /th/ ([ð]; [θ] – e.g.
"this"; "thin" respectively) but with a plosive sound instead of the natural fricative pronunciation (Warder: 2–3).
(g) For illustration purposes, it is worth mentioning that
consonants before the mentioned /th/ tend to become dentals naturally (e.g. letter
"d" in "rod thin").
(h) Thus, we can gather that this class of letters does not
represent the regular alveolars as articulated with the tongue
on the alveolar ridge – i.e. the noticeable ridge between just behind the upper teeth and the hard palate – as with /t/,
/d/ etc. (i) In the following, the Pāḷi dentals with equivalents and approximations in English:

• ta [t̪] = hit this.
• tha [t̪ʰ] ~ attack, with stronger breath pulse and the tongue in dental position.
• da [d̪] = rod thin.
• dha [d̪ʰ] ~ den, with breath pulse as with tha and the tongue in dental position.
• na [n̪] = tenth.
• la [l] = wea ̪ lth.
• sa [s] = salt.

Labials (oṭṭhaja)⁴³

(a) The letters of this group are u,
ū, pa, pha, ba, bha, ma and spoken in contact with both lips (Ñāṇadhaja,
2011: 12) – the end-group nasal is,
in addition to that, also produced in the nose (nāsikaṭṭhānaja; Rūp: 2);

⁴³ Lit.
"lip-born."

however, the lips should be open when articulating u,
^ū (Ñāṇadhaja: 14).
(b) The unvoiced and voiced letters form this class are pa,
pha and u, ū, ba, bha, ma respectively (Kacc 9; Duroiselle,
1906/1997: 4; Vidyabhusana, S. & Punnananda,
1935: 5).
(c) The lips should have strong contact when forming the labials,
excepting the vowels (Ñāṇadhaja: 14).

• u [u] = put.
• ū [uː] = fruit.
• pa [p] = stop.
• pha [pʰ] ~ prawn, with stronger breath pulse.
• ba [b] = hub.
• bha [bʰ] ~ hub, with breath pulse as with pha.
• ma [m] = moon.

Gutturo-palatal (kaṇṭhatāluja)

• (a) The letter is e, and its articulation happens in
  the throat (as with all other vowels) and the palate (Ñāṇadhaja,
  2011: 12), with open lips (Rūp: 2).
  (b) It was stated already that all vowels are voiced in the Pāḷi language (Duroiselle,
  1906/1997: 4).
• e [ɛ] = fell.
• e [eː] = Seele (German).⁴⁴

Gutturo-labial (kaṇṭhoṭṭhaja)

• (a) The letter is o and is produced in the throat (as with all other vowels) and the lips,
  with an effort to keep the lips open (Rūp: 2; Ñāṇadhaja,
  2011: 12, 14).
  (b) Again, all vowels need to be voiced in the Pāḷi language (Duroiselle,
  1906/1997: 4).
• [o] = oko (Czech).⁴⁵
• [oː] ~ home (hoʊm; corresponding to [o] before the sound change to [ʊ]).

⁴⁴ I am not aware of any American English equivalent.

⁴⁵ See previous footnote.

Dento-labial (dantoṭṭhaja)

(a) The letter is va and is generated with the teeth and the lips (Ñāṇadhaja,
2011: 12), voicing occurs (Kacc 9; Vidyabhusana,
S. & Punnananda, 1935: 5).
(b) It is formed by somewhat (tholaṃ) touching,
with slight contact (Rūp: 2; Ñāṇadhaja: 14).
(c) Duroiselle (1906/1997: 6) and Warder (1963/2001: 3) mention
that when this letter is preceded or combined with another consonant (e.g. as in tvā),
it is instead pronounced as a pure labial.

• va [v] = vine.
• va [w] = wind.

The Pure Nasal (niggahītaṃ)

(a) This letter (aṃ) ⁴⁶ is called niggahītaṃ
or anunāsiko in Pāḷi47 (Kacc 8) and is listed by all Pāḷi grammarians as a consonant (Kacc 6; Rūp: 3; Deokar,
2008: 113).
(b) It is never of long duration when pronouncing it (Geiger,
1916/1956: 63) and is described as neither voiced nor unvoiced (Deokar,
2009: 179).
(c) Traditionally, it is held that a dot (bindu) after
a short vowel⁴⁸ or the dot itself represent the niggahītaṃ
– with the ṃ being a transliteration of the dot,
referring to some traditional orthographical representations of the phoneme (e.g. in the Sinhala script),
equal to the anusvāraḥ and anunāsikaḥ in Sanskrit49 (Mogg: 42; Rūp: 3; Deokar,
2008: 114, 116; cf.
"Anusvāra," n.d.).
(d) In fact, manuscripts vary greatly in representing this letter; thus,
we find, for example, variants for the word
"other" as aṃño, añño or for the word
"question" as paṃha, pañha and paṇha (Frankfurter,
1883: 7).
(e) In the different modern American Standard

⁴⁶ The letter a is, again,
just added for ease of pronunciation.

⁴⁷ In Pāḷi, there is no difference between the anunāsiko
and the niggahītaṃ; both can be used interchangeably.
This can be gathered from numerous passages where the anunāsiko stands for the niggahītaṃ.
To quote the Paramatthajotikā I (p.
63) as an example, relating that the anunāsiko,
there clearly representing the niggahītaṃ,
was inserted for metrical reasons: sabbattha sotthiṃ gacchantīti
[...] anunāsiko cettha gāthābandhasukhatthaṃ vuttoti veditabbo.

⁴⁸ If a long vowel falls before a niggahītaṃ,
it is shortened (
"Pali," n.d.).

⁴⁹ It depends, however, which definition one refers to (cf.
Deokar, 2009: 180).

Code for Information Interchange (ASCII) transliteration schemes,
it is represented in a number of alternative ways – this,
nevertheless, makes no difference whatsoever as to pronunciation.
In the following, a selection of the more popular schemes:

• International Alphabet of Sanskrit Transliteration (IAST): ṃ.50
• International Organization for Standardization (ISO; ISO 15919): ṁ.51
• Indian Languages Transliteration (ITRANS): M; N; .m.
• Velthuis: .m.
• (a) The niggahītaṃ is capable of forming homorganic nasals,
  i.e. the place of articulation when pronouncing the niggahītaṃ
  is assimilated to that of the end-group nasals in the Pāḷi alphabet,
  thereby being displaced by them, these and the niggahītaṃ thus becoming distinct phonetically (Kacc 31,
  466, 537; Rūp: 28; cf.
  Deokar, 2008: 113; Levman, 2020: 29; Thitzana,
  2016: 52).
  (b) These are the respective parameters for the assimilation of the niggahītaṃ to the end-group nasals:
• ṅ before a velar stop (k, kh,
  g, gh).
• ñ before a palatal stop (ca, cha,
  ja, jha).
• ṇ before a retroflex stop (ṭa,
  ṭha, ḍa, ḍha).
• n before a dental stop (ta, tha,
  da, dha).
• m before a labial stop (pa, pha,
  ba, bha).
• (a) The place of articulation in the case of the niggahītaṃ is the nose (nāsikaṭṭhānaja –
  "born in the place of the nose" or nāsikaja –
  "nose-born"; Rūp: 2; Ñāṇadhaja, 2011: 12).
  (b) The place for the pronunciation of the end-group nasals
  is twofold: They are verbalized in the place of the nose (nāsikaṭṭhānaja),
  as is the niggahītaṃ, but in addition to that also in their
  "own place" of articulation (sakaṭṭhānaja),
  as elaborated above.
  (c) For example, the labial [m] is verbalized in the place
  of the nose (nāsikaṭṭhānaja) and with the lips (oṭṭhaja –
  "lip-born"; Rūp: 2; Ñāṇadhaja: 12).
  (d) This

⁵⁰ As in the Romanized editions of the Chaṭṭhasaṅgāyana (Sixth Buddhist Coun-

cil) and also in those of the later Pali Text Society.
51 This rendition also corresponds to the Unicode character.

demonstrates that the end-group nasals and the niggahītaṃ are distinct from each other.

(a) The Padarūpasiddhi mentions that this phoneme is called
niggahītaṃ because the instrument of articulation (karaṇaṃ) is restrained (niggahīta,
past passive participle of niggaṇhāti –
"press," "repress") by an obstructed opening (mukhenāvivaṭena)
and because it is based upon (nissāya) the short vowels a,
i, u, taking them up (gayhati,
passive form of gaṇhāti – also
"seize," "acquire," "grasp"; Rūp: 3; Deokar,
2008: 113; Warder, 1963/2001: 4).
(b) In fact, the 17th century CE Kaccāyana-vaṇṇanā (as cited by Deokar,
2009: 178, 181, n.
6) explains that the nose is defined as the articulatory organ
(nāsikāsaṅkhātaṃ karaṇaṃ) – and that this is what is restrained when pronouncing the niggahītaṃ.
(c) This process of restraining, as per that text,
also refers to the phrase
"by an obstructed opening" (mukhenāvivaṭena).
(d) This must refer to the partial occlusion instanced by the soft palate,
restricting parts of the open airflow of an released (vimutta)
or oral vowel and allowing parts to escape through the nose and the mouth respectively,
since a full occlusion would produce end-group nasals,
as discussed above (e.g. [m] when closing the lips or [ŋ] with occlusion of the oral cavity's back by the tongue).
(e) As a technical term, vimutta stands in direct opposition
to the niggahītaṃ and denotes denasalization (Deokar,
2008: 118).
(f) This must inform our understanding of the commentarial contrast of oral vowels with the niggahītaṃ:
"'Released (vimutta)': that which,
not having nasalized, is spoken with an unobstructed opening,
having let loose, [and] unrestrained (aniggahetvā) the articulatory organs [...]
'pattakallaṃ' should be spoken having nasalized with obstructed opening,
[but] 'pattakallā' is called an utterance freed (vimuttavacanaṃ) from the niggahītaṃ [i.e.
'the restrained'],
having not nasalized by an unobstructed opening
"52 (Sp V: 56).
On this contrast, see further also Deokar (pp.
112–3, 116–8).

⁵² Vimuttanti yaṃ karaṇāni aniggahetvā vissajjetvā vivaṭena mukhena anunāsikaṃ akatvā vuccati.
[...] pattakallan'ti avivaṭena mukhena anunāsikaṃ katvā vattabbe
'pattakallā'ti vivaṭena mukhena anunāsikaṃ akatvā vacanaṃ niggahitassa vimuttavacanaṃ nāma.

• (a) Scholars who investigated the phonetic reality of the
  niggahītaṃ now also seem to regard it as a nasalization of the short vowels a,
  i, u (Collins, 2006: 1; Deokar,
  personal communication, July 3, 2020; Geiger,
  1916/1956: 61; Levman, personal communication,
  July 8, 2020).
  (b) Therefore, the most likely pronunciation of the niggahītaṃ corresponds to nasalized vowels:
• aṃ [ã] = French (Québécois): "maman
  "
• iṃ [ĩ] = Portuguese: "sinto"
• uṃ [ũ] = Portuguese: "dum."
• (c) That this phenomenon of vowel nasalization is the correct
  interpretation is furthermore corroborated by the probability
  of it not having been a foreign element in the major autochthonous
  language groups present during the floruit of the Buddha.
  (d) These groups are the ancestral prototypes of modern languages
  in which this is a recognized feature (as in Dravidian Tamil or Santali).
  (e) In a similar way, this holds true for nasal assimilation (see above).
• (a) What emerges from the above is that the pronunciation
  of the niggahītaṃ as it is commonly realized in the traditional
  Buddhist countries (in Sri Lanka and Thailand as [ŋ] and in Burma as [m]) does not reflect early usage.
  (b) However, to dismiss these articulations offhand would be too rash an approach,
  as can be gleaned from the instructions on what the proper methods
  for giving the Three Refuges (tisaraṇaṃ) within the bounds
  of the going forth as a sāmaṇera (sāmaṇerapabbajjā) and the ordination as a bhikkhu (upsampadā) imply:
  "But with the giving these 'buddhaṃ saraṇaṃ gacchāmi
  ' as being joined as one (ekasambandhāni),
  they should be given having nasalized the end [of each word];'
  buddhaṃ saraṇaṃ gacchāmi,' having broken up [the words] (vicchinditvā),
  they should be given, having produced the letter m
  " (Sp III: 10).
  (c) Within the field of monastic legislation (vinaya),
  these two methods represent different but equally valid approaches
  to giving the Three Refuges during the mentioned procedures.53 (d) However,

⁵³ From a linguistic standpoint,
however, articulating an [m] at the end of a Middle Indic (MI,
incl.
Pāḷi) word does not constitute regular pronunciation.
According to the operant rules of MI,
it changes into the niggahītaṃ.

it is important to note that fastidious attention and adherence
to proper pronunciation – apart from the just-provided instance
of leeway – is an essential part of Theravāda legislative speech (kammavācā),
insofar that an invalidation happens if
"instead of an unaspirated sound [e.g. ka,
pa, ba] an aspirated sound [kha,
pha, bha], instead of an aspirate an unaspirate one,
instead of an oral one (vimutta) a niggahītaṃ,
instead of a niggahītaṃ an oral one is produced [...]
"54 (Sp V: 55–6).
(e) So, the statement of Deokar (2009: 181) that
"convention always supersedes grammar
" might be a good principle to adopt for monastic activities such as chanting but less so for legislative speech.

⁵⁴ [...] sithile kattabbe dhanitaṃ,
dhanite kattabbe sithilaṃ, vimutte kattabbe niggahitaṃ,
niggahite kattabbe vimuttanti imāni cattāri byañjanāni antokammavācāya kammaṃ dūsenti. However,
other confusions of letters such that of long ā for short a do not upset (kuppati) legislative speech.
There is also a difference of opinion between the suttanta
and vinaya elders in that the former group accepts a confusion of da for ta,
ta for da, ca for ja, ja for ca,
ya for ka and ka for ya, but for legislative speech this is not suitable,
whence the latter group avoids these faults (Sp V: 55–6).

Sandhi

• (a) Sandhi denotes the process of euphonic (or
  "pleasing," "harmonious") changes that may occur when two letters
  meet during the formation of words and compound words (Duroiselle,
  1906/1997: 6; Thitzana, 2016: 35; Perniola,
  1997: 7) and may be translated as
  "euphonic combination," "union," "junction
  " or "connection."
  (b) The changes transpire by means of elisions,
  substitution, augmentation etc. (Yindee,
  2018: 46).
  (c) When it occurs within a word,
  it is called akkharasandhi (
  "sandhi of letters," aka "internal sandhi
  ") and padasandhi ("sandhi of words,"
  aka "external sandhi") when two words meet,
  either to form compound words or to remain separate (Gair & Karunatillake,
  1998: 9; Nhu Lien, 2007: 28) – examples for internal sandhi:
  saṃ + yuttaṃ → saññuttaṃ; and external sandhi: tassa + iti → tassāti.
  (d) There are four types of sandhi:
• Vowel sandhi (sarasandhi): meeting of two vowels,
  as final and initial member.
• Consonantal sandhi (byañjanasandhi): meeting of final vowel and initial consonant.
• Niggahīta sandhi: meeting of the niggahītaṃ (ṃ) as final member and vowel or consonant as following initial.
• Natural sandhi (pakatisandhi): retention of the structural pattern with no union taking place.
• (a) The rules for the blending of two consonants also belong to the category of sandhi (Duroiselle,
  1906/1997: 6) but will be dealt with here in the following chapter
  "Morphology."
  (b) Let it be remarked that all the necessary morphological
  procedures to be elucidated in this chapter are on occasions applicable for the verbs of all roots (Kacc 517).
  (c) The references to Kaccāyana are as indicated after the rules,
  others are positioned next to the section headings.
  (d) Guide to the following section on sandhi:
• → = "becomes," "changes into," "results in.
  "
• → ∅ = elision.

• ∅ → = insertion.

• – / =
  "in the environment of."

• • = meeting.

• = word boundary.

  In this section

  • Vowel Sandhi (sarasandhi) (Duroiselle,
  • Exceptions:
  • Transformation into Semi-Vowels (ādeso) 56
  • Exceptions:
  • Exceptions:
  • Consonantal Insertion (āgamo)
  • Consonantal Sandhi (byañjanasandhi)
  • Niggahīta Sandhi (Duroiselle,
  • Natural Sandhi (pakatisandhi)
• [] = optionality (only after symbols).
• (V̆) = short vowel.
• (V̄) = long vowel.
• (C) = consonant.
• (CC) = double consonant.

(a) The underscores (__) indicate the position in the environment
where the action happens that is expressed as a rule before the slash; for example,
the formula:
"vowel → (V̄) [usually] / __ vowel [same class]
" says that any vowel (vowel) in the environment before another
vowel (/ __ vowel) of the same class is usually lengthened (→ (V̄) [usually]).
(b) If it should express that the lengthening would happen after (instead of before) another vowel,
one would simply change the element
"/ __ vowel" as above to "/ vowel __,"
with the underscores in the posterior position.
(c) If there is some additional rule after a comma,
following the element which occurs after the slash,
that indicates that it applies to this element when the change
of the pre-slash rule has occurred or simultaneously (e.g.
"vowel → ∅ [occasionally] / __ vowel,
vowel → (V̄)
" means after the vowel has been elided – which happens occasionally – when coming before another vowel,
that last-mentioned vowel is also lengthened (vowel → (V̄)).
(d) To give two other general examples to facilitate comprehension:55
"vowel → ∅ [usually] / o __" signifies that a vowel is usually elided in the environment after the vowel o.
Formula "v → b / # __" means that v changes into b after the beginning of a word – in the following,
the respective rules in full.

Vowel Sandhi (sarasandhi) (Duroiselle,

1906/1997: 7–12)

1. Vowel → ∅ / __ vowel (e.g. yassa + indriyāni → yassindriyāni; Kacc 12).

⁵⁵ For exemplification of explicit instances,
see just below.

• 1. Vowel / __ dissimilar vowel, dissimilar vowel → ∅ (e.g. cakkhu + indriyaṃ → cakkhundriyaṃ; Kacc 13).
• 1. Vowel → ∅ [occasionally] / __ vowel,
     vowel → (V̄) (e.g. tatra + ayaṃ → tatrāyaṃ; Kacc 15).
• 1. Vowel (V̆) → (V̄) [occasionally] / __ vowel,
     vowel → ∅ (e.g. vi + atimānenti → vītimānenti; Kacc 16).
• 1. a or ā → ∅ [occasionally] / __ i or ī,
     i or ī → e (e.g. upa + ikkhati → upekkhati; Kacc 14).
• 1. a or ā → ∅ [occasionally]/ __ u or ū,
     u or ū → o (e.g. canda + udayo → candodayo; Kacc 14).

Exceptions:

• a → (V̄) / __ iti, i → ∅ (e.g. tassa + iti → tassāti).
• a / __ i, i → ∅ (e.g. pana + ime → paname).
• ā → ∅ / __ i (e.g. seyyathā + idaṃ → seyyathidaṃ).
• 1. Vowel (V, V̆ ̄) → (V̄) [usually] / __ vowel [same class]
     (e.g. a + a → ā; i + ī → ī; ū + u → ū).
• 1. Vowels before particles beginning with a,
     i, e (e.g. atha, iva, eva) follow the rules of sandhi thus:
• itthī + iti → itthīti.
• e / __ e, e → ∅ (e.g. sabbe + eva → sabbeva).
• o → ∅ / __ e (e.g. so + eva → sveva).
• a → ∅ / __ ettha (e.g. na + ettha → nettha).
• 1. e → ∅ / __ dissimilar (V̄) (e.g. me + āsi → māsi).
• 1. e → ∅ / __ dissimilar (V̆) followed by (CC) (e.g. sace + assa → sacassa).
• 1. Vowel → ∅ [usually] / o __ (e.g. cattāro + ime → cattārome).
• 1. Vowel (V̄) → (V̆) [occasionally] / __ eva,
     eva → ri (e.g. yathā + eva → yathariva; Kacc 22).
• 1. abhi → abbh / __ dissimilar vowel (e.g. abhi +
     uggacchati → abbha uggacchati → abbhugacchati; Kacc 44,
     46).

• 14. ti → c [occasionally],
  c → cc (e.g. iti + etaṃ → iccetaṃ; Kacc 19,
  28, 47).

• 1. di → jj / __ dissimilar vowel (e.g. nadī + ā → najjā).
• 1. adhi → ajjha / __ dissimilar vowel (e.g. adhi + okāse → ajjhokāse; Kacc 45).

Transformation into Semi-Vowels (ādeso) 56

• 1. i → y / __ dissimilar vowel (e.g. vi + ākāsi → vyākāsi; Kacc 21).
• 1. e [of me, te, ke, ye etc.] → y /__a followed by (CC) (e.g. ke + assa → kyassa).
• 1. e [of me, te, ke, ye etc.] → y / __ a followed by (C),
     a → (V̄) (e.g. me + ahaṃ → myāhaṃ: cf.
     Kacc 17).

Exceptions:

• e → ∅ / __ vowel (V̄) (e.g. me + āsi → māsi).
• e → ∅ / __ vowel (V̆) followed by (CC) (e.g. sace + assa → sacassa).
• e / __ vowel, vowel → ∅ (e.g. te + ime → teme).
• e → ∅ / __ a → (V̄) (e.g. sace + ayaṃ → sacāyaṃ).
• 1. u → v [occasionally] / __ dissimilar vowel (e.g. anu + eti → anveti; Kacc 18).
• 1. o → v [occasionally] / __ dissimilar vowel (e.g. yo + ayaṃ → yvāyaṃ; Kacc 18).

Exceptions:

• u → ∅ / __ dissimilar vowel(e.g. sametu + āyasmā → ametāyasmā).
• u → (V̄) / __ i (e.g. sādhu + iti → sādhūti).
• o → ∅ [usually] / __ vowel (V̄) followed by (CC).
• o → ∅ / __ vowel (V̆) followed by (CC) (e.g. kuto + ettha → kutettha).

⁵⁶ This term also refers to the transformation of one vowel into another (Thitzana,
2016: 570).

Consonantal Insertion (āgamo)

• 1. To avoid a hiatus, not seldom the following letters are inserted between two vowels: y,
     v, m, d, n, t, r, l (= ḷ),
     h (e.g. na + imassa → nayimassa; √bhū + ādāya → bhūvādāya; idha + āhu → idhamāhu etc.; Kacc 35).
• 1. Vowel → ∅ / __ consonant, ∅ → o [occasionally] (e.g. para + sahassaṃ → parosahassaṃ; Kacc 36).
• 1. Vowel / __ vowel or consonant,
     ∅ → ṃ (e.g. ava + siro → avamsiro; Kacc 37).
• 1. Putha, ∅ → g [occasionally] / __ vowel (e.g. putha + eva → puthageva; Kacc 42).
• 1. ā [of pā] → (V̆), ∅ → g [occasionally] / __ vowel (e.g. pā + eva → pageva; Kacc 43).

Consonantal Sandhi (byañjanasandhi)

(Duroiselle, 1906/1997: 15–6)

• 1. Vowel (V̄) → (V̆) [occasionally] / __ consonant (e.g.
     yiṭṭhaṃ vā hutaṃ vā loke → yiṭṭhaṃ va hutaṃ va loke; Kacc 26).
• 1. Vowel (V̆) → (V̄) / __ consonant (e.g. √su + rakkhaṃ → sūrakkhaṃ).
• 1. Vowel (V̆) / __ consonant, (C) → (CC) (e.g. idha + pamādo → idhappamādo; usually after: u,
     upa, pari, ati, pa, a, anu,
     etc.).
• 1. Vowel (V̆) → (V̄) [occasionally] / __ consonant (e.g. muni + care → munī care; Kacc 25).
• 1. Vowel → ∅ and is replaced by a [occasionally] / __ consonant (e.g. eso dhammo → esa dhammo; Kacc 27).
• 1. Vowel → bb / __ v (e.g. ni + vānaṃ → nibbānaṃ).
• 1. Vowel / __ consonant, consonant (C) → (CC) (e.g. idha pamādo → idhappamādo; Kacc 28).
• 1. Vowel (V̄) [of particles] → (V̆) [usually] / __ reduplicated consonant (e.g. ā + kamati → akkamati).

============================================

• 1. o [of so, eso, ayo, mano,
     tamo, paro, tapo and a few others] → a [occasionally]
     / __ consonant (e.g. esa dhammo; sa attho; ayapattaṃ).
• 1. ava → o [occasionally] / __ consonant (e.g. ava + naddha → onaddha; Kacc 50).
• 1. dha → da [occasionally] / __ vowel (e.g. ekaṃ + idha + ahaṃ → ekamidāhaṃ; Kacc 20).
• 1. dha → ha [occasionally] (e.g. rudhira → ruhira; Kacc 20).
• 1. d → t (e.g. sugado → sugato; Kacc 20).
• 1. t → ṭ (e.g. pahato → pahaṭo; Kacc 20).
• 1. t → k (e.g. niyato → niyako; Kacc 20).
• 1. t → dh (e.g. gantabba → gandhabbo; Kacc 20).
• 1. tt → tr (e.g. attajo → atrajo; Kacc 20).
• 1. tt → cc (e.g. batto → bacco; Kacc 20).
• 1. g → k (e.g. hatthupaga → hatthupaka; Kacc 20).
• 1. r → l (e.g. paripanno → palipanno; Kacc 20).
• 1. y → j (gavayo → gavajo).
• 1. y → k (e.g. saye → sake; Kacc 20).
• 1. vv → bb (e.g. kuvvato → kubbato; Kacc 20).
• 1. k → y (sake pure → saye pure).
• 1. j → y (nijaṃputtaṃ → niyaṃputtaṃ; Kacc 20).
• 1. k → kh (nikamati → nikhamati; Kacc 20).
• 1. p → ph (e.g. nipatti → niphatti; Kacc 20).
• 1. pati → paṭi [occasionally] / __ vowel (Kacc 48).
• 1. putha [inter alia] → puthu / __ consonant (Kacc 49).

Niggahīta Sandhi (Duroiselle,

1906/1997: 14–5)

• 1. ṃ / __ consonant (e.g. taṃ dhammaṃ kataṃ).
• 1. ṃ → respective nasal: ṅ, ñ,
     ṇ, n, m [occasionally] / __ consonant (e.g. raṇaṃ
• jaho → ranañjaho; taṇhaṃ + karo → taṇhaṅkaro; saṃ + ṭhito → saṇṭhito; Kacc 31).
• 1. ṃ → l / __ l (e.g. paṭi + saṃ + līno → paṭisallīno; saṃ + lekko → sallekho).
• 1. ṃ → ñ [occasionally] / __ e [or h] (e.g. taṃ
• eva → taññeva; evaṃ + hi → evañhi; Kacc 32; for doubling of the consonant,
  see under
  "Consonantal Sandhi (byañjanasandhi),"
  pt.
  7.; Kacc 28).
• 1. ṃ → ñ [occasionally] / __ y (e.
     g.
     saṃ + yuttaṃ → saññuttaṃ; Kacc 33; for doubling of the consonant,
     see under
     "Consonantal Sandhi (byañjanasandhi),"
     pt.
     7.; Kacc 28).
• 1. ṃ → d [occasionally] / __ vowel (e.g. etaṃ + attho → etadattho; Kacc 34).
• 1. ṃ → m [occasionally] / __ vowel (e.g. taṃ ahaṃ → tamahaṃ; Kacc 34).
• 1. ṃ → ∅ [occasionally] / __ consonant (e.g. ariyasaccānaṃ + dassanaṃ → ariyasaccānadassanaṃ; Kacc 39).
• 1. ṃ → ∅ [occasionally] / __ vowel (e.g. tāsaṃ + ahaṃ santike → tāsāhaṃ santike; Kacc 38).
• 1. Vowel → ∅ [occasionally] / ṃ __ (e.g. kiṃ + iti → kinti; Kacc 40).
• 1. Vowel → ∅, consonant (CC) → consonant (C) / ṃ __ (e.g. evaṃ assa → evaṃsa; Kacc 41).
• 1. ∅ → ṃ / __ vowel [or consonant] (e.g. ava siro → avaṃsiro; Kacc 37).

Natural Sandhi (pakatisandhi)

• 1. Vowel / __ consonant (e.g. mano + pubbaṅgamā → manopubbaṅgamā; Kacc 23).
• 1. Vowel / __ vowel (e.g. ko imaṃ; Kacc 24).
• 1. i [and u] / __ any verb w/ vowel initial (e.g. gāthāhi ajjhabhāsi).
• 1. i [and u] / __ any verb.
• 1. Vowel / __ vocative case (e.g. kassapa etaṃ).
• 1. Final long vowel remains unchanged if not followed by iti or not being compounded.
• 1. Vowel / __ particle w/ initials other than a,
     i, e (e.g. atha kho āyasmā).

Morphology

(a) The following morphological changes happen mostly in the formation of the passive,
past passive participle, the stems built from the third class root affixes,
of the infinitive, absolutive, the future passive participle
and in the formation of the desiderative – also under the influence of certain affixes in the derivation of nouns.
⁵⁷ (b) Regressive assimilation (←) is the more common.
(c) The ṇ placed traditionally before all causative affixes
to denote vowel increase (vuddhi) in the root (see below the chapter
"Vowel Gradation") is always to be elided (e.g. √kara + ṇaya + ti → kārayati; Kacc 523).
(d) References to Kaccāyana are again as indicated – others are placed next to the section headings.
(e) The paradigms are as follows (Duroiselle,
1906/1997: 17–8):

• 1. Mute58 → mute / mute __ (e.g. √saja + ta → satta).
• 1. Dental → guttural / guttural __ (e.g. √laga + na → lagna → lagga).
• 1. Dental voiceless → retroflex / palatals __ (e.g. √maja
• ta → maṭṭha or maṭṭa); j and c → t [occasionally]
  / __ t (e.g. √bhuja + ta → bhutta; √muca + ta → mutta).
• 1. Dental voiceless → retroflex / retroflex __ (e.g. √kuṭṭa + ta → kuṭṭha).
• 1. Dental → consonant / __ consonant (e.g. √uda + gaṇhāti → uggaṇhāti).
• 1. Voiced aspirate → voiced unaspirate / __ t,
     t → dh (e.g. √rudhi + ta → rud + dha → ruddha).
• 1. Voiceless unaspirated guttural or labial → voiceless dental
     / __ voiceless dental (e.g. √tapa + ta → tatta).

⁵⁷ See chapters "Kita and Taddhita Affixes
" and "Uṇādi Affixes."

⁵⁸ Mute because they require closure or contact (phasso)
in their place of articulation and the stopping of the breath.
Not to be confused with surd; i.e. unvoiced consonants.
They are: k, kh, g, gh, c, ch,
j, jh, ṭ, ṭh, ḍ, ḍh, t,
th, d, dh, p, ph, b, bb.
As with the letters in the alphabet, the a appended to the Pāḷi roots provided is just for ease of utterance.

• 1. Voiced or voiceless unaspirated dental → labial / __ labial (e.g. tad + purisa → tappurisa).
• 1. n → ∅ [occasionally ṃ] / __ ta (of past passive participle; e.g. √mana + ta → mata).

Assimilation of y⁵⁹ (Duroiselle,

1906/1997: 19–20)

• 1. Consonant ← y (e.g. √divu + ya → divva → dibba),
     also in the middle of a compound word (e.g. api + ekacce → apyekacce → appekacce).
• 1. vv → bb (e.g. √divu + ya → divva → dibba).
• 1. ā [of √dā, √pā, √hā, √mā and √ñā] → eyya [occasionally] / __ ya (e.g. √dā + ya → deyyaṃ –
     "something to give"; Kacc 544).
• 1. ∅ → ya [occasionally] / da and dha-ending roots __ tuna,
     tvāna and tvā [suffixes] (e.g. u + pada + ya + tvā → uppajjitvā; Kacc 606).
• 1. ty → cc (e.g. √sata + ya → satya → sacca).
• 1. dy → jj (only after √mada and √vada; e.g. √mada + ya → madya → majja; Kacc 544).
• 1. dhy → jjh (e.g. √rudha + ya → rudhya → rujjha).
• 1. thy → cch (e.g. tath + ya → tathya → taccha).
• 1. my → mma (Kacc 544).
• 1. jy → gga (e.g. √yuja + ya → yogga; Kacc 544).
• 1. y → sibilant / sibilant __ (e.g. √pasa + ya → pasya → passa).
• 1. v → b / # __ (e.g. vi + ākaraṇaṃ → vyākaraṇaṃ → byākaraṇaṃ).
• 1. Dental → y / __ y (e.g. √ud + yuñjati → uyyuñjati).
• 1. u → (V̄) [of √guha and √dusa] / __ causative affixes (e.g. √guha + aya + ti → gūhayati –
     "causes to protect," "hide"; Kacc 486).
• 1. ya → abba / √bhū __ (e.g. √bhū + ya → bhabbo; Kacc 543).
• 1. a and v [of √vaca, √vasa, √vaha] → u [occasionally]
     / __ ya (e.g. √vaca + ya + ti → vuccati; Kacc 487).

⁵⁹ Assimilation of this type happens mostly in the formation of the passive voice,
absolutives, verbal bases/stems of the third class and derived nouns.

• 1. Initial vowels [of √dā, √dhā,
     √mā, √ṭhā, √hā, √pā, √maha,
     √matha] → ī / __ ya (e.g. dā + ya + ti → dīyati; Kacc 502).
• 1. Consonant y [of √yaja] → i / __ ya (e.g. √yaja + ya + ti → ijjate –
     "He is worshipped"; Kacc 503).

Assimilation of r (Duroiselle,

1906/1997: 21)

• 1. r → ∅ / __ mute (e.g. √kara + ta → kata).
• 1. r → ∅, a → (V̄) / __ mute (incl.
     lengthening of preceding a; e.g. √kara + tabba → kātabba).
• 1. n → ṇ/ r __, r → ṇ (e.g. √cara + na → carṇa → ciṇṇa).
• 1. r → l / __ l (e.g. dur + labho + si [o] → dullabho).
• 1. When any r-morpheme is appended to a root,
     the first component vowel of that root and its last consonant
     are usually elided as well the vowel and the r of the r-morpheme (Kacc 539).

Assimilation of s (Duroiselle,

1906/1997: 21–2)

• 1. j + sa → kkh / __ sa (e.g. bubhuj + sa → bubhukkha).
• 1. p + sa → cch / __ sa (e.g. jigup + sa → jiguccha).
• 1. t + sa → cch / __ sa (e.g. tikit + sa → tikiccha).
• 1. s + sa → cch / __ sa (e.g. jighas + sa → jighaccha).
• 1. y → s [occasionally] / sa __ (e.g. √nasa + ya → nassa; alasa + ya + si [aṃ] → ālasyaṃ).
• 1. t → ṭ / s __ (e.g. √kasa + ta → kaṭṭha).
• 1. Dental → s / __ s (e.g. √uda + sāha → ussāha).
• 1. s → t / __ t (e.g. √jhasa + ta → jhatta).
• 1. s → tth / __ t (e.g. √vasa + ta → vuttha).

Assimilation of h (Duroiselle,

1906/1997: 22–3)

• 1. Consonant → aspirated consonant / __ h (e.g. √uda + harati → uddharati).
• 1. hṇ → ṇh/ __ ṇ (e.g. √gaha + ṇa → gahṇa → gaṇha).⁶⁰

⁶⁰ Kacc 490 explains it like this: h [of √gaha] → ∅ when ∅ → ṇhā (e.g. gaṇhāti).

• 1. h y and in some instances ya → la (e.g. oruh + ya → oruyha; Kacc 488).
• 1. h v (e.g. jihvā → jivhā).
• 1. h → y [seldom]/ __ y (e.g. leh + ya → leyya).
• 1. h → gh [occasionally] / # __ (e.g. hammati → ghammati).
• 1. h + t → ddh (e.g. √duha + ta → duddha).
• 1. h + t → dh (sometimes; e.g. √liha + tuṃ → ledhuṃ).

Reduplication⁶¹ (Duroiselle,

1906/1997: 83)

• 1. The second and fourth consonants of the consonant groups (sing.
     vaggo) are added to the first and third respectively (e.g. yatra ṭhitaṃ → yatraṭṭhitaṃ; Kacc 29).62
• 1. Initial vowel [of roots] → (V̄) (e.g. √ah → āha).
• 1. The reduplicated vowels → i, ī and a [occasionally] (e.g. jigucchati; Kacc 465).
• 1. A guttural is reduplicated by its corresponding palatal (e.g. √kita + cha + ti → cikicchati; Kacc 462).
• 1. An unaspirate is always reduplicated by an unaspirate (e.g. √chida → ciccheda –
     "It was cut"; Kacc 458, 462).
• 1. An aspirate is reduplicated by its unaspirate (e.g. √bhuja → bubhukkhati; Kacc 458,
     461).
• 1. The initial h of a root is reduplicated by j (e.g. √hā → jahāti; Kacc 464).
• 1. v is reduplicated by u [usually] (e.g. √vasa → uvāsa).
• 1. a or ā takes a (e.g. √dhā → dadhā; Kacc 460).
• 1. i or ī takes i [occasionally] (e.g. √kita → cikicchā; Kacc 460).
• 1. u or ū takes u but occasionally a (e.g. √bhū → babhuva).
• 1. i → e [occasionally] (e.g. √chida → ciccheda).
• 1. u → o [occasionally] (e.g. √suca → susoca).

⁶¹ Mainly doubling of the first root consonant together
with a following vowel in the formation of the active base/stem of the third class of conjugation.

⁶² Exceptions: idha, cetaso,
daḷhaṃ, gaṇhāti, thāmasā.

• 1. a [of a root] → (V̄) (e.g. √vada → uvāda).
• 1. m [of √māna] → v [occasionally] / __ reduplicated vowel (e.g. vīmaṃsati; Kacc 463).
• 1. √māna → maṃ [occasionally] / reduplicated vowel __ sa (e.g. vīmaṃsati; Kacc 467).
• 1. √pā → vā [occasionally] / reduplicated vowel __ sa (e.g. pivāsati; Kacc 467).
• 1. Reduplicated k [of √kita] → t / __ reduplicated vowel (e.g. tikicchati; Kacc 463).
• 1. ∅ → ṃ [occasionally] / reduplicated vowel __ (e.g. caṅkamati; Kacc 466).

Further Morphological Changes

• 1. √pā → pivā [occasionally] (e.g. √pā + ā + ti → pivati; Kacc 469).
• 1. √ñā → jā, jaṃ, nā [occasionally] (e.g. √ñā + a + ti → jānāti; Kacc 470).
• 1. √disa → passa, dissa, dakkha [occasionally] (e.g. √disa + a + ti → passati; Kacc 471).
• 1. √hara → gī / __ sa (e.g. jigīsati; Kacc 474).
• 1. √brū and √bhū change into āha and bhūva respectively
     / __ perfect endings (e.g. √brū + a → āha; Kacc 475).
• 1. m [of √gamu] → cch [occasionally] / __ all conjugational
     root affixes (e.g. √gamu + a + māna + si [o] → gacchamāno; Kacc 476).
• 1. Initial vowel [of √vaca] → o / __ aorist suffix (e.g. √vaca + uṃ → avocuṃ; Kacc 477).
• 1. ū [of √hū] → eha, oha, e [occasionally] / __ future tense suffix,
     future tense suffix may → ∅ (e.g. √hū + ssati → hehiti; Kacc 480).
• 1. √kara may → kāha [occasionally] / __ future tense suffix,
     future suffix → ∅ (Kacc 481).
• 1. ā [of √dā] → aṃ / __ present tense suffixes mi
     and ma (e.g. √dā + mi → dammi; Kacc 482; ṃ → m by Kacc 31).
• 1. Non-conjunct root vowels → increase63 [usually] / __
     noncausative affixes (e.g. √hū + a + ti → hoti; Kacc 485).
• 1. ∅ → kha, cha, sa [occasionally] / √tija,
     √gupa, √kita and √māna __ (e.g. √tija + kha + ti → titikkhati –
     "He forbears [or 'endures']"; Kacc 433).
• 1. √gaha → ghe / __ affix ppa (e.g. gheppati; Kacc 489).
• 1. √kara → kāsa [occasionally] / __ aorist suffix (e.g. √kara + ī → akāsi; Kacc 491).
• 1. Suffix mi → mhi [occasionally] / √asa __ (e.g. √asa + mi → amhi –
     "I am"; Kacc 492).
• 1. Suffix ma → mha [occasionally] / √asa __ (e.g. √asa + ma → amha –
     "We are"; Kacc 492).
• 1. Suffix tha → ttha [occasionally] / √asa __,
     s [of √asa] → ∅ (e.g. √asa + tha → attha –
     "You are"; Kacc 493).
• 1. Suffix ti → tthi [occasionally] / √asa __ (e.g. √asa + ti → atthi –
     "[there] is"; Kacc 494).
• 1. Suffix ti → ssa / √asa __ (e.g. √asa + ti → assa
     "It should be"; Kacc 571).
• 1. ∅ → i / √brū __ ti (e.g. √brū + a + ti → bravīti
     "He says"; Kacc 520).
• 1. Suffix tu → tthu [occasionally] / √asa __ (e.g. √asa + tu → atthu –
     "Let it be"; Kacc 495).
• 1. s of [of √asa] → ∅ when nominative suffix si⁶⁴
     is appended to √asa (e.g. √asa + si → asi –
     "You are"; Kacc 496).
• 1. Aorist suffixes ī → ttha / √labha __ (e.g. √labha + ī → alattha; Kacc 497)
• 1. iṃ → tthaṃ / √labha __ (e.g. √labha + iṃ → alatthaṃ; Kacc 497).
• 1. Aorist suffix ī → cchi / √kusa __,
     s [of √kusa] → ∅ (e.g. √kusa+ ī → akkocchi –
     "He reviled"; Kacc 498).
• 1. √dā → dajja [occasionally] (e.g. √dā + eyya → dajjeyya; Kacc 499).

⁶³ See below the chapter
"Vowel Gradation" for details.

⁶⁴ This nominative suffix undergoes changes to o,
aṃ etc. in other cases.

• 1. √vada → vajja [occasionally] (e.g. √vada + eyya → vajjeyya; Kacc 500).
• 1. √gamu → ghamma [occasionally] (e.g. √gamu + a + tu → ghammatu –
     "Let him go"; Kacc 501).
• 1. Aorist suffix uṃ → iṃsu / all roots __ (Kacc 504).
• 1. √jara → jīra or jiyya [occasionally] (e.g. √jara + a + ti → jīrati; Kacc 505).
• 1. √mara → miyya [occasionally] (e.g. √mara + a + ti → miyyati; Kacc 505).
• 1. Initial vowel a [of √asa] → ∅ [occasionally] / __
     all suffixes (e.g. √asa + a + anti → santi; Kacc 506).
• 1. √asa → bhū [occasionally] (e.g. √asa + a + ssanti → bhavissanti; Kacc 507).
• 1. Optative suffix eyya → iyā or ñā / √ñā __ (Kacc 508).
• 1. Affix nā (fifth class active base root affix) → ∅ or ya [occasionally] / √ñā __ (Kacc 509).
• 1. Affix a (first class active base root affix) → ∅ or e
     [occasionally] (e.g. √vasa + a + mi → vademi; Kacc 510).
• 1. Affix o (seventh class active base root affix) → u
     [occasionally] / √kara __ (e.g. √kara + o + te → karume –
     "He does"; Kacc 511).
• 1. Component vowel a [of √kara] → u [occasionally] (e.g. √kara + o + ti → kurute –
     "He does"; Kacc 511, 512).
• 1. The increase morpheme o → ava / √bhū,
     √cu etc. __ vowel (e.g. √cu + a + ti → cavati; Kacc 513).65
• 1. The increase morpheme e → aya / √nī,
     √ji etc. __ vowel (e.g. √ji + a + ti → jayati; Kacc 514).
• 1. Increase vowel o → āva, e → āya / __ causative affix [e,
     ya] (e.g. √lū + e + ti → lāveti; Kacc 515).
• 1. ∅ → i / root consonant __ asabbadhātuka suffixes66 (e.g. √gamu + ssati → gamissati; Kacc 516).

⁶⁵ See also below the chapter
"Vowel Gradation" for details.

⁶⁶ Suffixes of the perfect (parokkhā),
aorist (ajjatanī), future indicative (bhavissanti) and conditional (kālātipatti) are meant (Kusalagñāṇa,
2012: 161).

• 1. Last component vowel [of polysyllabic roots] → ∅ [occasionally] (e.g. √mara + a + ti → marati; Kacc 521).
• 1. Consonants s and m [of √isu,
     √yamu] → cch [occasionally] (e.g. √isu + a + ti → icchati; Kacc 522).
• 1. ima → a, samāna → sa / ima,
     samāna, apara __ suffixes jja,
     jju, jja, jju (e.g. ima + jja → ajja –
     "today").
• 1. Kita affix ta → cca or ṭṭa / √naṭa __ (e.g. √naṭa + ta + si [aṃ] → naccaṃ –
     "dancing"; Kacc 571).
• 1. Regarding kita affix ta:
• i.
  √sāsa, √disa → riṭṭha / __ ta (e.g. √disa + ta → diṭṭha –
  "seen"; Kacc 572).
• ii.
  ta → ṭṭha [together with final root consonant] / √puccha,
  √bhanja, √hansa and roots ending in s etc. __ (e.g. √bhanja + ta → bhaṭṭha; Kacc 573).
• iii.
  ta → uṭṭha [together with final s of the root] / √vasa __,
  v → u [occasionally] (e.g. √vasa + ta → vuṭṭha; Kacc 574–575).
• iv.
  ta → dha and ḍha respectively / dha,
  ḍha, bha, ha __ (e.g. √budha + ta + si [o] → buddho –
  "the Awakened One"; Kacc 576).
• v.
  ta → gga [together with final j of the root] / √bhanja __ (e.g. √bhanja + ta → bhagga –
  "broken"; Kacc 577).
• vi.
  ta → (CC) / √bhanja etc. __, final root consonant → ∅ (e.g. √caja + ta → catta –
  "given up": Kacc 578).
• vii.
  ta → (CC) / √vaca __, v [of √vaca] → u [occasionally],
  c → ∅ (e.g. √vaca + ta → utta –
  "said"; Kacc 579).
• viii.
  ta → (CC) / √vaca __, v [of √vaca] → u [occasionally],
  c → ∅, ∅ → v (e.g. √vaca + ta → vutta –
  "said"; Kacc 579).
• ix.
  ta → (CC) / √gupa etc. __, final root consonsants → ∅ (e.g. √lipa + ta → litta –
  "annointed"; Kacc 580).

• x.
  ta → iṇṇa / √tara etc. __, final root consonants → ∅ (e.g. saṃ + √pūra + ta → sampuṇṇa –
  "well filled"; Kacc 581).

• xi.
  ta → inna, anna, īṇa / √bhida etc. __,
  final root consonants → ∅ (e.g. √bhida + ta → bhinna –
  "broken"; Kacc 582).

• xii.
  ta → nta [occasioanlly] / prefix pa etc. + √kamu etc. __,
  final root consonants → ∅ (e.g. pa + √kamu + ta → pakkanta; Kacc 584).
• xiii.
  ta → kkha and kka / √susa, √paca,
  √saka etc. __, final root consonants → ∅ (e.g. √susa + ta → sukkha –
  "dried"; Kacc 583).
• xiv.
  ta → ha / ha-ending roots (except √daha and √naha) __,
  h [of the roots] → ḷ (e.g. √baha + ta → bāḷha –
  "grown"; Kacc 589).
• 1. Initial a [of √yaja] → i / __ ṭṭha (morphological
     resultant of ta; e.g. √yaja + ta → yiṭṭha; Kacc 610; see also pt.
     ii above for changes which result in ṭṭha).
• 1. Final consonants [of ha, da, bha of √naha,
     √kudha, √yudha, √sidha, √labha,
     √rabbha etc.] → da / __ dha (morphological resultant of ta; e.g. √labha + ta → laddha –
     "obtained"; Kacc 611; see also pt.
     iv above for changes which result in dha).
• 1. Final component consonants ha,
     ḍha [of √daha, √vaḍha] → ḍa / __ ḍha (morphological resultant of ta; e.g. √daha + ta → daḍḍha –
     "burnt"; Kacc 612; see also pt.
     iv above for changes which result in ḍha).
• 1. Regarding kita affixes ta and ti:
• i.
  Initial vowel [of √jana] → ā / __ ta or ti (e.g. √jana + ta → jāta –
  "born," "arisen"; Kacc 585).
• ii.
  Final root consonant [of √gamu, √khanu,
  √hana, √ramu etc.] → ∅ [occasionally]/ __ ta or ti (e.g. √khanu + ti → khati –
  "digging"; Kacc 586).
  Exception: ∅ → i as per pt.
  67 below (Kacc 617).

• iii.
  Final r [of √kara, √sara etc.] → ∅ / __ ta or ti (e.g. pa + √kara __ ti → pakati –
  "original [or 'natural'] form"; Kacc 587).

• iv.
  Vowel ā [of √ṭhā, √pā etc.] → i or ī respectively / __ ta or ti (e.g. √pā + ti → pīti –
  "act of drinking"; Kacc 588).

• 1. ta [of kita affix tabba] → raṭṭha / √sāsa,
     √disa etc. __ (e.g. √disa + tabba + si [aṃ] → daṭṭhabbaṃ; Kacc 572,
     elision of r according to Kacc 539).
• 1. tuṃ suffix → raṭṭhum / √sāsa,
     √disa etc. __ (e.g. √disa + tuṃ = daṭṭhuṃ; Kacc 573; elision of r according to Kacc 539).
• 1. Regarding kita affix ṇa:
• i.
  nja [of √ranja] → j / __ ṇa (Kacc 590).
• ii.
  √hana → ghāta / __ ṇa (e.g. go + √hana + aka + si [o] → goghātako –
  "the one who kills cows"; Kacc 591).
• iii.
  √hana → vadha / __ ṇa (e.g. √hana + ṇa + si [o] → vadho
  "the one who kills"; Kacc 592).
• iv.
  vowel ā [of ā-ending roots] → āya / __ ṇa (e.g. √dā + aka + si [o] → dāyako –
  "a donor"; Kacc 593).
• 1. √kara → kha / pura, saṃ,
     upa and pari __ (e.g. saṃ + √kara + ta → saṅkhata –
     "conditioned," "prepared"; Kacc 594).
• 1. √kara → kā / __ kita suffixes tave and tuna (e.g. √kara + tuna → kātuna –
     "having done"; Kacc 595).
• 1. m and n [of √gamu, √khanu,
     √hana etc.] → n [occasionally] / __ kita affixes tuṃ
     and tabba (e.g. √gamu + tabba + si [aṃ] → gantabbaṃ –
     "that which should be done"; Kacc 596).
• 1. Kita suffixes tuna, tvāna,
     tvā etc.:
• i.
  → ya [occasionally] / after all roots __ (e.g. ā + √dā + tvā → ādāya; Kacc 597).
• ii.
  → racca [occasionally] / all ca and na-ending roots __ (e.g. vi + √vica + tvā → vivicca –
  "having renounced," "being far from"; Kacc 598).

• iii.
  → svāna, svā [occasionally] / √disa __ (e.g. ā + √disa + tvā → disvā; Kacc 599).

• iv.
  → mma, yha, jja, bbha, ddha [occasionally] ma-,
  ha-, da-, bhaending roots __ (e.g. ā + √gamu + tvā → āgamma –
  "having come"; Kacc 599).

• 1. ∅ → i / root __ all affixes (ririya,
     tabba, ta, tvā etc.; e.g. √vida + tabba → viditabba; Kacc 605).
• 1. The first n [of some roots] → ṃ (e.g. √ranja + ṇa + si [o] → raṅgo –
     "act of coloring"; Kacc 607).
• 1. √ge → gī [whenever appropriate] (e.g. √ge + ta + si [aṃ] → gītaṃ –
     "music"; Kacc 608).
• 1. √sada → sīdā [always] (e.g. ni + √sada + a + ti → nisīdati; Kacc 609).
• 1. √gaha → ghara [occasionally] / __ affix ṇa (e.g. √gaha + ṇa + si [aṃ] → gharaṃ –
     "house"; Kacc 613).
• 1. da [of √daha] → ḷa [occasionally] / __ affix ṇa
     (e.g. pari + √daha + ṇa + si [o] → pariḷāho –
     "burning"; Kacc 614).
• 1. Final consonant [of a root] → ∅ / __ kita affix kvi (i.e. other roots themselves; Kacc 615).
• 1. ∅ → ū / √vida __ kita affix kvi (e.g. lokavidū
     "the knower of the world"; Kacc 616).
• 1. (a) When an inserted i (as per Kacc 605) is already positioned,
     the final consonants [of √hana, √gamu,
     √ramu, √saka, √kara etc.] are not elided with ta affixes.
     (b) Applicable affixes are: tabba,
     tuṃ, tvā and tvāna.
     (c) Inapplicable exceptions are: tave,
     tāye, tavantu, tāvi and teyya (Kacc 617; Thitzana,
     2016: 756).
• 1. r [of √kara] → t / __ tu (e.g. √kara + ritu + si [→ ∅] → kattā –
     "the one who does"; Kacc 619).
• 1. r [of √kara] → t [occasionally] / __ tuṃ,
     tuna, tabba (e.g. √kara + tuna → kattuna; Kacc 620).
• 1. The final component consonant c [of √paca etc.] and
     j [of √yaja etc.] → k and g respectively / __ affix
     ṇa (e.g. √yuja + ṇa + si [o] → yogo; Kacc 623) but not / __ ṇvu affixes (Kacc 618).

Uṇādi Rules

• 1. Initial vowel [of √gaha] → ge [occasionally] (e.g. √gaha + a + si [aṃ] → gehaṃ –
     "house"; Kacc 629).
• 1. su [of stem masu] → cchara or cchera (e.g. masu + kvi + si [o] → maccharo –
     "jealousy"; Kacc 630).
• 1. √cara → cchariya, cchara or cchera / ā __,
     ā → (V̆) (e.g. ā + √cara + kvi + si [aṃ] → accharaṃ; Kacc 631).
• 1. tha [of √matha] → la (e.g. √matha + a + si [o] → mallo –
     "wrestler"; Kacc 634).
• 1. Some roots which end in c and j → k and g respectively
     / __ ṇ-initial affix (e.g. √sica + ṇa + si [o] → seko –
     "pouring"; Kacc 640).
• 1. una [of stem suna "dog"] → oṇa,
     vāna, uvāna, ūna, unakha, una,
     ā or āna (Kacc 647).
• 1. Stem taruṇa → susu (Kacc 648).
• 1. uva [of stem yuva] → uvāna,
     una or ūna (e.g. yūno –
     "youth"; Kacc 649).
• 1. ū, u and asa [of √sū, √vu,
     √asa] → ata, ∅ → affix tha (√sū + tha + si [aṃ] → satthaṃ –
     "a weapon"; Kacc 660).
• 1. √hi → heraṇ or hīraṇ / paṭi __ (e.g. paṭi +
     √hi + kvi + si [aṃ] → pāṭihīraṃ or pāṭiheraṃ –
     "miracle"; Kacc 662).
• 1. Stem putha → puthu, patha,
     ∅ → affix ama⁶⁷ [occasionally] (e.g. putha + kvi [→ ∅] + si [→ ∅] → pathavī –
     "earth"; Kacc 666).

⁶⁷ For an example, refer to the section
"Ordinal Numerals."

Vowel Gradation

(a) Root vowels may vary in "strength
" or appear in various "grades," which means that they are changed into another vowel sound.
(b) This process is called "strengthening
" or "vowel gradation" and occurs regularly in the formation of verbal stems,
non-finite verbs (i.e. infinitives and absolutives) and in the
derivation of words while appending certain affixes (see chapters
"Kita and Taddhita Affixes" and
"Uṇādi Affixes"; Dhammajoti, 2018: 20; Duroiselle,
1906/1997: 23).
(c) Three grades exist in Pāḷi, named and following a generic paradigm as illustrated in Table 1 below.

Table 1. Vowel Gradation

(a) The ancient grammarians explain these processes as an absence and prefixing (or
"increase") of the letter a respectively (Dhammajoti,
2018: 20; Warder, 1963/2001: 12) or as the replacement of i and u by e and o respectively (Kacc 401).
(b) These processes can also be reversed; an instance of such is called viparīto (
"reversal"; Thitzana, 2016: 569).
(c) Examples of vowel gradation can be conveniently given with the words bhūta (unstrengthened form),
bhavati and hoti (strengthened form),
bhāveti (increased form), which are all formed from √bhū.
(d) The increase morpheme o → ava,
e → aya/ √bhū, √cu etc. /__ vowel (e.g. √cu + a + ti → cavati; Kacc 513–514).
(e) Vowel o → āva, e → āya / __ causative affix [e,
ya] (e.
g.
√lū⁶⁸ + e + ti → lāveti; Kacc 515).

⁶⁸ The ū of √lū has previously been strengthened to o.

Parts of Speech (padajāti)

• 1. Nouns incl.
     adjectives and pronouns (nāmāni).
• 1. Verbs (ākhyātāni).
• 1. Indeclinable prepositions and prefixes (upasaggā or upasārā).
• 1. Indeclinable particle conjunctions,
     prepositions, adverbs and all other indeclinables (nipātā).

Sentence Structure and Syntax

• (a) The main collections (sing.
  nikāyo) of Pāḷi Buddhist texts employ an idiom which usually bears a close affinity to the syntax of Vedic,
  thereby manifesting a closer linguistic connection to Indo-European than Classical Sanskrit; however,
  marked divergences from Vedic nevertheless exist (cf.
  Hendriksen, 1944: 81; Wijesekera, 1936/1993: 6).
  (b) In the Pāḷi language, the usual order within sentences is: subject (S; kattā – lit.
  "agent") + object (O; kammaṃ) + verb (V; kamma kiriyā; e.g. bhagavā (S) dhammaṃ (O) deseti (V) –
  "The Blessed One teaches the dhamma,"
  DN III: 41 [DN 28]), with the verb at the end of the sentence (Yindee,
  2018: 69).
  (c) An adjective normally precedes the noun it qualifies and a modifying adverb its verb.
  (d) "Adverbs of time always come first in the sentence
  " (Duroiselle, 1906/ 1997: 52).
• (a) A regular yet not universal feature of prose portions
  in the Pāḷi language (as well as Vedic and Buddhist Hybrid Sanskrit)
  is the grouping of word elements with related or identical meaning (e.g. synonyms),
  a remnant of the oral style of composition and transmission,
  facilitating memory (Allon, 1997: 191; Anālayo,
  2009: 740–1).
  (b) It is commonplace to encounter two,
  three or more adjectives qualifying the same substantive noun
  and to find a body of substantive nouns functioning as the sentence
  subjects or objects; the same principle of aggregation holds also true for verbs and adverbs (Allon: 191).
  (c) The arrangement of all these parts of speech is determined by the so-called Waxing Syllable Principle,
  influencing the sequential order according to the syllable

length of the word strings: each subsequent element stays either
equal in syllable length or increases; waning does not occur
(e.g. bhīto saṃviggo lomahaṭṭhajāto [syllable count: 2 + 3 + 6] –
"frightened, agitated and gotten horripilation,"
DN I: 24 [DN 2]; Anālayo: 740).
(d) Exceptions to this are groupings of items which must
belong together due to similar grammatical and morphological form,
related meaning and doctrinal content (e.g. regarding the items
of the noble eightfold path [ariyo aṭṭhaṅgiko maggo]; Allon: 191; Anālayo: 741).
(e) When a sentence has two subjects,
the last one determines the person for the corresponding verbs (Kacc 409).
(f) The enclitics me and te are,
as a rule, found in the second position of a sentence (Oberlies,
2019: 265).
(g) Clauses are conjoined in the following ways (Perniola,
1997: 384, 390–1):

• They simply follow each other.
• Relative clauses and phrases:
• With relative pronouns, adjectives or adverbs as the sentence initial of the subordinate clause,
  in correlation with a demonstrative pronoun,
  adjective or adverb introducing the main clause (e.g. yo dhammaṃ passati so buddhaṃ passati –
  "He who sees the dhamma is the one who sees the Buddha,"
  Mil: 35).
• With a participle functioning as an adjective,
  agreeing with a noun (e.g. addasā kho āyasmā rāhulo bhagavantaṃ dūratova āgacchantaṃ –
  "Ā.
  Rāhula saw the Blessed One, who was coming from afar,"
  MN II: 40 [MN 61]).
• With dependent-determinative, descriptive-determinative or attributive compounds⁶⁹ (e.g. evaṃ kho,
  kassapa, bhikkhu sīlasampanno [tappurisa compound] hoti –
  "thus, Kassapa, is a bhikkhu one who is possessed of virtue,"
  DN I: 81 [DN 8]).
• With the introduction of adverbs or adverbial phrases of time and space (e.g. tadā –
  "at that time"; tattha – "there"; bhūtapubbaṃ –

⁶⁹ See chapter "Compounds (samāsā)
" for details.

• "formerly"; ekaṃ samayaṃ "at one time
  "; tena samayena "at that time"; atha kho –
  "now then" etc.).
• With particles ca (copulative) and vā (disjunctive).
• Phrase kuto pana ("still less") and words pageva (
  "still more"), aññādatthu ("except
  "; all adversative).
• With seyyathāpi ("just as") contrasted with evameva (
  "just so") and yathā ("just as") contrasted with tathā (
  "so"; all comparatives).
• Consecutive and connected verbs may stand in the absolutive with the finite verb being placed last.
• (a) It may often happen that the verb
  "to be" is not expressed but only implicitly understood (e.g. rūpaṃ aniccaṃ –
  "Form is impermanent").
  (b) In the end, there are no hard and fast regulations about the sentence structure – the subject,
  to proffer an example, remains the subject even if it succeeds the object (e.g. dhammaṃ buddho [S] deseti –
  "Dhamma teaches the Enlightened One [S]
  "; Dhammajoti, 2018: 9, 12).
  (c) The word order is rarely of grammatical significance but in terms of style and emphasis rather relevant (Warder,
  1963/2001: 15).
  (d) The vocative is unable to stand as sentence initial (Palistudies,
  2018b).
  (e) Regarding case endings (see below),
  the Padarūpasiddhi (Rūp: 29) and Duroiselle (1906/1997: 24)
  mention that at times whole syllables are elided for the sake
  of meter or to facilitate simplicity (e.g. abhiññāya sacchikatvā → abhiññā sacchikatvā).

Nouns (nāmāni) 70

Kinds of Nouns (cf.

Thitzana, 2016: 184–5)

• 1. Substantive Nouns (nāmanāmāni).
• i.
  Common nouns (sādhārananāmāni).
• ii.
  Proper nouns (asādhārananāmāni).
• 1. Adjectives (guṇanāmāni).
• 1. Pronouns (sabbanāmāni).
• 1. Compound nouns (samāsanāmāni; Kacc 601).
• 1. Nouns formed from taddhita affixes (taddhitanāmāni,
     incl.
     numerical nouns; Kacc 601).
• 1. Nouns formed from kita affixes (kitanāmāni; Kacc 601).71

General Characteristics

• (a) In the Pāḷi language, there are no fundamentally distinct classes of substantive nouns,
  adjectives and pronouns, all being united under the broad category of nāmaṃ (noun),
  but individual differences nonetheless exist (Pandita,
  n.d.; Warder, 1963/2001: 7).
  (b) Substantive nouns and adjectives share declensions (see Table 3 in the
  "Tables" section) and are also part of the samāsanāmāni,
  taddhitanāmāni and kitanāmāni classes; e.g.:
• Sāriputto; arindamo (nāmanāmaṃ,
  kitanāmaṃ and samāsanāmaṃ).
• Kaccāyano (nāmanāmaṃ and taddhitanāmaṃ).⁷²
• (a) Although adjectives bear the name of guṇanāmaṃ (
  "quality noun") – indicating that they are a class of nouns qualifying
  other nouns – the lack of an absolute distinction between substantive
  nouns and adjectives can be seen in many instances; for example,
  the word kusala (
  "wholesome," "skillful") can stand as a substantive noun: kusalaṃ (
  "the wholesome")

⁷⁰ Sing.
nāmaṃ.

⁷¹ The last three-mentioned items are dealt with in separate chapters.

⁷² I am indebted to Ā.
Kovida (Myanmar, aka Sayadaw U Kovida) for initially clarifying
the concept for me and providing the examples (personal communication,
April 11, 2020).

or operate as an attribute of another noun,
as in kusalo dhammo (
"the good dhamma").
(b) Compound nouns are simply combinations made up of members
from the above-given noun classes (see the respective chapters for details).
(c) Although particles (sing.
nipāto) and prefixes (sing.
upasaggo or upasāraṃ) cannot be classified under the rubric
of nouns – possessing no gender and number – they can be subject
to the rules of nouns when standing as independent words in a sentence; these are,
however, exceptional cases (Thitzana,
2016: 185).
(d) It should be noted that the case endings in these unusual
instances are not permitted to remain attached to these words (Thitzana: 323).

General Formation

The formation of nouns in the Pāḷi language comes about in the following manner,
conjoining two or more of these elements:

• Prefix (upasaggo or upasāraṃ).
• Root (dhātu).
• Kita affix (kitapaccayo).
• Taddhita affix (taddhitapaccayo).
• Interfix (āgamo).
• Suffix (paccayo or vibhatti), expressing:
• Case.
• Number.
• Gender.
• (a) For example, the substantive noun āvāso is formed
  from these elements: ā (upasaggo) + √vas + a (kitapaccayo)
  form the stem to which si [o] (vibhatti; singular nominative case masculine suffix) is appended; thus,
  finally → āvāso (
  "home," "dwelling place").
  (b) Another example to illustrate how an interfix is applied is given with the following.
  The adjective mānasika is broken up like this: √māna + s (āgamo) + ika (taddhitapaccayo) → mānasika (
  "related to mind") or + si [aṃ] (singular nominative case neuter sufffix –
  "that which is related to mind") when functioning as a substantive noun.
  (c) Another interfix, consonant n,
  is

added in the formation of numerical nouns with dative suffix naṃ (e.g. dvinnaṃ –
"two"; Kacc 67).
(d) The rules of sandhi and morphology are regularly applied (i.e. assimilation,
elision, insertion and reduplication; e.g. √ana + ya [taddhitapaccayo] → añña [pronoun] –
"other," "another").⁷³

Gender, Number and Case

(a) In the Pāḷi language, three genders (sing.
liṅgaṃ) exist for nouns: masculine (pulliṅgaṃ),
feminine (itthiliṅgaṃ) and neuter (napuṃsakaliṅgaṃ; Kacc 52; Oberlies,
2019: 199); two numbers: singular (ekavacanaṃ) and plural⁷⁴ (bahuvacanaṃ; Buddhadatta,
1937/1997: 3; Warder, 1963/ 2001: 7); three persons (sing.
puriso): first (paṭhamapuriso), second (majjhimapuriso)
and third (uttamapuriso),⁷⁵ with the indigenous grammatical tradition recognizing seven cases.
Seven cases because it treats the vocative as standing apart from the rest (cf.
Yindee, 2018: 58–9), although in modern grammars,
we may find it included among the total count,
therewith amounting to eight cases (Collins,
2006: 18–9).
(b) Substantive nouns usually assume only one gender,
whereas adjectives and pronouns are capable of standing in all three – variably corresponding to the gender,
number and case of the nouns of which they are the attribute or correlate respectively.
(c) However, a substantive noun, without changing form,
can also possess two or all three genders (e.g. kumāra [
"boy," "prince"] can assume masculine [kumāro] as well as feminine forms [kumārī],
and

⁷³ See chapters "Sandhi,"
"Morphology" and "Kita and Taddhita Affixes
" for details and for explanations on the characteristics of roots and stems the section
"General Formation" in the chapter "Verbs (ākhyātāni).
"

⁷⁴ The plural has basically replaced the dual in Pāḷi; only a few forms can be

found (Oberlies, 2019: 201).
75 This is the schema of the traditional Pāḷi grammarians – first
and third persons given therein are the exact opposites in English (e.g. English first persons
"I" and "we" are each classed as third person [uttamapuriso] in Pāḷi,
and English third persons
"he/she/it" and "they" each correspond to the first person [paṭhamapuriso] in Pāḷi); however,
to forestall confusion, occurrances of grammatical numbers in
Pāḷi – within the bounds of the present grammar – correspond,
henceforth, to English usage.

kuddaka ["belly"] can occur in the masculine [kuddako],
neuter [kuddakaṃ] and feminine [kuddakā]; Geiger,
1916/1994: 67–8; Yindee: 58).

Substantive Nouns (nāmanāmāni)

As mentioned above, this classification includes common and proper nouns (cf.
Buddhadatta, 1937/1997: 48).

• Common nouns: a group of unspecified people (vāṇijo
  "merchant"), animals (hatthī – "elephant
  "), places (nagaraṃ – "city"), things (rukkho –
  "tree") and ideas (i.e. abstract nouns; dhammo –
  "norm," "nature").
• Proper nouns: specific persons (sāriputto right-hand chief disciple of Lord Buddha),
  places (rājagaho – an ancient Indian city with that name) and organizations.
• (a) As single entities, substantive nouns have usually merely one gender (of the three,
  as mentioned above), but as final members of attributive compounds,⁷⁶
  substantive nouns can also assume all three genders – in which case they are used adjectivally (Collins,
  2006: 17; Buddhadatta, 1937/1997: 4).
  (b) As mentioned in the previous section,
  a substantive noun may possess altogether two or even all three genders.

Adjectives (guṇanāmāni) 77

(a) As adverted to earlier, adjectives bear the name of guṇanāmāni (
"quality nouns"), indicating that they are a class of nouns modifying other nouns,
providing more information about them (Yindee,
2018: 55).
(b) Adjectives are capable of standing,
as mentioned above, in all three genders – assuming the gender,
number and case of the nouns of which they are the attributes.
They themselves lack intrinsic gender (Dhammajoti,
2018: 9).
(c) If an adjective qualifies two or more nouns,
it may agree with the sum (being plural then) or with the nearest.
(d)

⁷⁶ See chapter "Compounds (samāsā)
" for details.

⁷⁷ For the different stems of substantive nouns and adjectives as well as their declensions,
see Table 3 in the section
"Tables" at the end of this book.

The qualified words may also taken to be collective with the adjective in singular.
(e) In case of gender conflict, the masculine has precedence over the feminine and the neuter over both.
(f) They usually precede the qualified noun,
but in the case of several adjectives modifying a noun,
one adjective may precede and the rest follow.
(g) An adjective following a substantive noun usually indicates the predication of it,
with the attribute being emphasized and translated by a relative clause
"who/which is ..."
or an adjectival predicate (i.e. an adjective that follows a linking verb,
such as
"am," "is," "are," "has been"), modifying the sentence subject (e.g. vedanā aniccā –
"feeling is impermanent," MN I: 146 [MN 35]; cf.
Bodhi, 2020: 32– 3).
(h) Without a verb in the sentence (i.e. in equational sentences),
this predication may take place regardless of the adjective's positioning (cf.
Warder, 1963/2001: 60–1).
(i) The distinction between substantive nouns and adjectives is not an absolute,
as previously explained (Warder: 62).

(a) Pronouns or pronominal adjectives are used as adjectives (Ānandamaitreya,
1993/2012: 25).
(b) Almost all pronouns become adjectives when they are used before a substantive noun of the same gender,
number and case.
(c) They are pronouns when they stand alone in a sentence.
(d) Let it be reiterated briefly that stems in a are more common than any other (Geiger,
1916/1956: 159; Warder, 1963/2001: 8).
(e) The following affixes are used to form adjectives: a,
ana, ika, ka, kaṭa, ma, mantu (mā; possessive),
maya, min (mī; possessive), ra,
so, (s)sī (possessive), tana,
ti, tya, va, vantu (vā; possessive),
vī.
78

Three Grades of Adjectives

(a) To express the comparative form of adjectives,
the following affixes are appended to nominal bases: tara,
iya, iyya and for the superlative: tama,
iṭṭha, issika, (i)ma (Ānandamaitreya,
1993/2012: 93; Duroiselle, 1906/ 1997: 59; Perniola,
1997: 48–9; Yindee, 2018: 55–6).79 (b) In most cases,
affixes tara and tama may be used interchangeably as well as iya for iṭṭha,

⁷⁸ See chapters "Kita and Taddhita Affixes
" and "Uṇādi Affixes."

⁷⁹ See Table 2 below for a summary.

issaka and vice versa (Duroiselle: 59).
(c) Affix tara may be superadded to some forms of the comparative
and superlative without affecting the meaning as conveyed by the initial affix.
(d) Some adjectives have more than one form or are constructed irregularly (e.g. antika –
"near" → nediya – "nearer" or nediṭṭha –
"nearest"; Duroiselle: 60; Perniola: 49).
(e) The comparative adjectives are primarily employed with
nouns in the instrumental or ablative cases singular and the
superlatives with the genitive and locative cases plural (e.g. tayā mahiddhikataro –
"more powerful than you"; tesaṃ sattamo –
"the best of them"; tesu sattamo –
"best among them"; Ānandamaitreya: 93) – the adjective may also stand after the ablative case.
(f) The declensions of adjectives of stems in mantu (mā),
vantu (vā), vī are elided together with the antecedent vowel (e.g. guṇavā + iya → guṇiyo –
"more virtuous").
(g) Depending on which gender they ought to take,
the comparative and superlative forms are to be declined like
neuter and masculine stems in a or feminine stems in ā (Duroiselle: 59–60).
(h) Certain past participles, by themselves or as the final member of compounds,
are used comparatively with the appropriate affixes (e.g. paṇīta –
"excellent" → paṇītatara – "more excellent
").
(i) Within comparative compounds,
the first member commonly is the object of the comparison (e.g. paṇītasaṅkhatatara –
"reckoned as more excellent").
(j) Some substantive nouns can also take the comparative affixes (e.g. malatara –
"having a greater stain") and certain pronouns,
prepositions and adverbs are able to take both comparative and superlative affixes (e.g. aññatama –
"one out of many"; upari – "above" → uparima –
"uppermost"; Perniola: 50–1).

Table 2. Comparative and Superlative Degree of Adjectives

Note: Substantive nouns in nt take a before tara and tama,
forming the alternative stem in anta.
Sources: (a) Ānandamaitreya, B. (2012).
Pali made easy.
Buddhist Cultural Centre (original work published 1993).
https://archive.org/details/
PaliMadeEasyOCRed; (b) Perniola,
V. (1997).
Pali grammar.
The Pali Text Society.

Participles

The participles have the nature of verbal adjectives and must,
therefore, agree with the nouns they qualify in number,
gender and case (Duroiselle, 1906/1997: 100; Oberlies: 571,
f.n.
1; Perniola, 1997: 357).

Possessive Adjectives⁸⁰

Formation. (a) Commonly added are vantu (vā),
vī (Perniola, 1997: 147).
(b) The usual affixes that form possessives include: a,
āla, ālu, ava, ita, ika, imā,
in (ī) (strengthening as a rule takes place),
iya, ila, eyya, eyyaka, mantu (mā),
(s)sī, vantu (vā), vī.
(c) The feminine of possessive adjectives formed with mantu (mā) and vantu (vā) is usually,
but not invariably, constructed from the weakened base (Collins,
2006: 58).

Usage. (a) The possessive adjectives can be rendered into
English as regular adjectives or in combination with such words and idioms as
"having," "possessed of," "possessing
" (e.g. satimā – "possessed of

⁸⁰ For the declensions of the adjectives and substantive nouns in in (ī),
mantu (mā), vantu (vā), vī,
see Table 3 in the
"Tables" section and for further information on these as well as the remaining affixes,
the chapter
"Kita and Taddhita Affixes."

mindfulness [i.e. 'mindful']"; Kacc 369).
(b) The possessive adjectives in mantu (mā),
vantu (vā) and in (ī) are also used as substantive nouns
when they don't qualify any other noun81 (e.g. dhanavanto –
"the wealthy one"; gomī – "cattle,"
"a possessor of cattle"; Ānandamaitreya,
1993/2012: 53, 57; Duroiselle, 1906/1997: 150).
(c) As is the case with other adjectives,
possessive adjectives are capable of being rendered as relative
clauses and participle clauses/constructions with present participles (e.g. [relative clause] daṇḍī –
"the one who possesses [or 'carries'] a staff
"; [participle construction] medhāvī –
"the one possessing wisdom"; Kacc 364,
532; cf.
Hendriksen, 1944: 7, f.n.
2).

Adjectives from Pronominal Bases

(a) Duroiselle, 1906/1997: 77–8: Affixes di (dī),
disa, dikkha, disaka, risa (
"intimate," "likeness," "resemblance
") are, with lengthening of the base vowel,
appended to pronominal bases to form adjectives (e.g. ta + di → tādi –
"like him," "like that," "such").
(b) For that purpose, affixes tara,
tama (see above) are also added to the bases of interrogative pronouns,
but the meaning differs just a little from that of the respective bases (e.g. katara –
"which [of two]?"; katama – "which [of many]
").

Pronouns or Pronominal Adjectives (sabbanāmāni)

Kinds of Pronouns

• 1. Personal pronoun (puggalanāmaṃ).
• 1. Demonstrative pronoun (nidassananāmaṃ).
• 1. Relative pronoun (anvayīnāmaṃ).
• 1. Interrogative pronoun (pucchānāmaṃ).
• 1. Indefinite pronoun (anīyamanāmaṃ).
• 1. Possessive pronoun (Collins, 2006: 61; Nwe Soe,
     2016: 205; Perniola, 1997: 52).

⁸¹ Most, presumably all, of those formed with the other affixes too.

General Characteristics

(a) Substantive nouns and adjectives may qualify their referent words,
but pronouns act as mere pointers to these (Collins,
2006: 62).
(b) They are officially regarded as belonging to the category of nouns (nāmāni),
as mentioned above, and do not constitute a separate class of words.
(c) Pronouns can stand in all genders,
as they can represent all things and persons of the different genders apiece (Thitzana,
2016: 287) but are not able to be in the vocative case.
(d) We can find three persons for pronouns,
singular and plural.
(e) Almost all pronouns become adjectives when they are used before a noun of the same gender,
number and case (e.g. so puriso –
"that man").
(f) They are pronouns when they stand alone in a sentence (Buddhadatta,
1937/1997: 49–50).
(g) Pronouns can also act as noun substitutes (e.g. sā bhuñjati –
"She eats"; Collins: 62; Warder, 1963/2001: 7) and are oftentimes merely implied (e.g. bhuñjati –
"He eats").
(h) As a means of showing respect,
especially (but not exclusively) to persons of higher rank,
it is possible that the first and second person plurals of pronouns
may replace the respective singular forms; this usage is known as the
"honorific plural" (e.g. [possibly] yo amhākaṃ rājaputto vā rājamahāmatto –
"the one who is our prince and prime minister [...],"
Mil: 44).
(i) The second person singular, on the other hand,
is used for peers and subordinates,
as with German
"du" (Collins: 65).

General Formation⁸²

(a) For a description on the general features of the formation process of nouns (incl.
pronouns), see the above section of the present chapter having the same name as this one (i.e.
"General Formation"), with some additional specifics in the following.
(b) The i and a vowels of pronouns may lengthen when in certain combinations with √disa,
so too then vowel i of √disa (e.g. ya + √disa + kvi → yādiso –
"any kind of person"; Kacc 642).
It may be that the d of √disa → r,
and its s → sa, kkha or ī in the process

⁸² For a part of different pronouns and their declensions,
see Table 4 in the
"Tables" section as well as the following.

(e.g. yādikkho; Kacc 642).
(c) In the formation of the nominative case,
all pronouns may take interfix consonant k (e.g. sabbako –
"all"; Kacc 178).

The Traditional Inventory of 27 Pronouns (sabbanāmāni)

(1) sabba ("all"), (2) katara ("which [of two]?
"), (3) katama ("which [of many]?"),
(4) ubbaya (
"both"), (5) itara ("other [of two]
"), (6) añña ("other [of many]"), (7) aññatara (
"other [of many]"), (8) aññatama (
"a certain [of two]"), (9) pubba (
"former"), (10) para ("another"), (11) apara (
"another"), (12) dakkhiṇa ("right,"
"south"), (13) uttara ("upper,"
"north," "more than"), (14) adhara (
"lower"), (15) ya ("who," "what"), (16) ta (
"he," "that"), (17) eta ("this"), (18) ima (
"this"), (19) amu ("that"), (20) kiṃ (
"what?", "why?"), (21) eka ("one"),
(22) ubha (
"both"), (23) dvi ("two"), (24) ti (
"three"), (25) catu ("four"), (26) tumha (
"you"), (27) amha ("I," "we") (Sadd I: 150; Collins,
2006: 61).

Personal Pronouns

Usage. (a) Personal pronouns of the first and second persons
do not possess gender and invariably operate as substantive noun substitutes (Collins,
2006: 62).
(b) Enclitic genitive, dative, instrumental singular me (
"by me," "my," "mine") and genitive, dative,
instrumental and ablative plural no (
"our," "for us," "by us," "from us") are never used at the beginning of a sentence (Duroiselle,
1906/1997: 68), the same holds true for te.
(c) For enhanced emphasis, the personal pronouns are found with relative pronoun yo (e.g. lābhā vata me,
suladdhaṃ vata me, yohaṃ evarūpehi sabrahmacārīhi saddhiṃ viharāmī –
"It is indeed a gain for me, a great gain for me,
that I dwell together with such fellow companions in the holy life,"
MN I: 135 [MN 31]; Duroiselle: 74).

Demonstrative Pronouns

Usage. (a) The pronouns of absence,
formed from the stem ta(d), are employed to refer to someone
or something previously mentioned in a narrative or to absent
persons or things.⁸³ (b) They might serve as

⁸³ Pronoun ena is used in the same way (Oberlies,
2019: 277).

a connecting element between two sentences or to extend paragraphs or sections and be translated as
"that" etc. (e.g. etthantare yaṃ caritaṃ,
sabbaṃ taṃ bodhipācanaṃ –
"What conduct there is in the interim,
all that is a maturing of enlightenment,"
Cp: 1; yamahaṃ jānāmi, taṃ tvaṃ jānāsi –
"What I know is that what you know!",
DN I: 41 [DN 3]).
(c) They may also function as personal pronouns
"he," "she," "it" and other forms of those.
(d) The forms with n substitute (e.g. enaṃ) ⁸⁴ usually refer to a noun which was already mentioned.
(e) Besides these usages, forms of ta(d) can also be
employed adverbially in all oblique cases (e.g. taṃ [acc.] –
"thereto," "therefore," "that is why,"
"now," "then"; tamenaṃ – "at once
"; tassa [gen.] – "therefore"; yena [...] tena [instr.] –
"where [...], there," "here and there
"; tasmā [abl.] – "for this reason,"
"therefore"; tato – "from there,"
"thence"; tahiṃ [loc.] – "there,"
"therefore"; tahaṃ – "there"; Oberlies,
2019: 260, 270, f.n.
1; PED, s.v.
"ta").
(f) They are also used pleonastically,
giving emphasis (e.g. sohaṃ –
"I" [lit.
"that I"]).⁸⁵

• (a) Demonstrative pronouns formed from pronominal stem
  eta(d) are used to point to someone or something present
  in direct speech or to what immediately precedes or follows – they may be translated as
  "this" etc. (Collins, 2006: 63; Oberlies,
  2019: 274; Warder, 1963/2001: 29); they may also operate as the personal pronouns
  "he," "she," "it" etc. (b) Demonstrative pronouns with stems
  in ta(d)/eta(d) are often used to give slight emphasis
  to the subject or as something approximating the English article
  "the."
• (a) Demonstrative pronouns formed from the pronominal stem
  in ima (such as ayaṃ) are used similarly but convey a special sense of proximity or immediacy,
  whereas those constructed from eta(d) are merely indefinite (Oberlies,
  2019: 284; Warder, 1963/2001: 30).
  (b) They may be rendered into English with
  "this (here)" etc. (c) For enhanced emphasis,
  the pronouns ayaṃ and so/eso are applied pleonastically
  with relative yo (e.g. tasmā yoyaṃ me attapaṭilābho –
  "therefore, this is the acquisition

⁸⁴ These are given in Table 4 in the
"Tables" section.

⁸⁵ For the usage with a relative pronoun giving emphasis,
see below.

of personality for me," Sv-pṭ: 135; Duroiselle,
1906/1997: 74).
(d) The pronouns formed from amu imply a sense of being far away and may be translated as
"yonder" etc. (Collins, 2006: 63; Warder: 30).

Relative Pronouns

Formation.
(a) Relative pronouns are mainly found building relative clauses (e.g. yo dhammaṃ passati,
so buddhaṃ passati –
"He who sees the dhamma is the one who sees the Buddha,"
Mil: 35), but some are employed as indeclinables (Warder,
1963/2001: 70).
(b) Neuter singular of ya(d) is used preceding vowels and in compounds (Duroiselle,
1906/1997: 74).
(c) The relative pronouns take their gender and number from
the antecedent (the modified element from the main clause) in relative clauses,
but case is determined according to their function in the relative clause (Collins,
2006: 68).
(d) The relative clause regularly precedes the main clause (Warder: 71),
though it may be that the correlating pronoun (with its clause)
precedes the relative one (e.g. amataṃ tesaṃ bhikkhave sacchikataṃ,
yesaṃ kāyagatāsati sacchikatā –
"Those, bhikkhus, that have realized the deathless are those who realized mindfulness directed to the body,"
AN I: 29 [AN 1.627]; Dhammajoti, 2018: 103).

Usage. (a) Relative pronouns are commonly translated with
"who" or "which," in the three genders.
(b) As a simple marker of a relative clause or a connector of a subordinate clause,
it may function as an indeclinable and be translated as
"that," "since," "if," "whereas" etc. (e.g. nesa dhamma,
mahārāja, yaṃ tvaṃ gaccheyya ekako –
"It is not right, great king, that you might go alone,"
Jā II: 188 [Jā 547]; Dhammajoti, 2018: 102).
(c) For the sake of emphasis, the relative pronoun yo is
used pleonastically with demonstrative pronouns ayaṃ and so
as well as with personal pronouns (or eso; e.g. yo so,
āvuso, bhikkhu kāmesu avītarāgo hoti [...],
tasssa cittaṃ na namati ātappāya [...] – lit.
"That bhikkhu, friend, who has not been without passion regarding sensuality,
his mind does not bend toward ardor,"
DN III: 105 [DN 33]; Duroiselle, 1906/1997: 74).
(d) To express the sense of generalization,
the pronoun ya(d) can be employed thus:

• Repetition of ya(d) and the correlative in a distributive sense (e.g. yo yo [...] ādiyissati,
  tassa tassa dhanamanuppadassāmi –
  "Whoever will take up, to him I will give,"
  DN III: 27 [DN 26]).
• In combination with its correlative (e.g. yasmiṃ tasmiṃ
  "in whatever place/case").
• In combination with the indefinite pronouns (e.g. yaṃ kiñci –
  "whatever").
• (a) The form yadidaṃ can be employed in a variety of ways (e.g.
  "that is to say," "since," "which is this,"
  "namely"; Collins, 2006: 69).
  (b) The instrumental form yena in combination with a verb of motion carries the import of
  "where," "toward" and governs the nominative case (Warder,
  1963/2001: 73); in that sense, constructions with yena [...] tena may be translated as
  "where [...], there" (e.g. yena bhagavā tenupasaṅkami –
  "Where the Blessed One was, there [he] went,"
  Vin I: 4 [Pār 1]).
  (c) The relative pronouns in all oblique cases are also used adverbially (e.g. yena [instr.] –
  "where" or "at which place"; yena yena [instr.] –
  "wherever"; yahiṃ [loc.] – "where" or
  "whither"; yasmā [abl.] – "because
  " [corresponding with tasmā]; Oberlies,
  2019: 260, f.n.
  1, 270; PED, s.v.
  "ya").

Interrogative Pronouns

Formation and Usage.(a) Interrogative pronouns are used to formulate questions (Collins,
2006: 67).
(b) There is no equivalent to a question mark in the Pāḷi language; however,
if a sentence contains an interrogative pronoun,
the whole sentence takes on an interrogative sense.
(c) It is also possible that a sentence contains no such
pronoun but that the context imposes an interrogative sense (Warder,
1963/2001: 73–4).
(d) The particles su/ssu, nu and no⁸⁶ are used with interrogative pronouns for emphasis (PED,
s.v.
"ka"; Perniola, 1997: 56).
(e) As mentioned in the

⁸⁶ The last two-mentioned particles are used to signify doubt,
by themselves or with an interrogative pronoun or adverb.
They are oftentimes followed by kho (e.g. ko nu kho,
bho gotama, hetu ko paccayo [...] –
"Now what is the reason, dear Gotama,
what the cause [...]?
", DN I: 68 [DN 5]; Oberlies, 2019: 282,
f.n.
2).

section on adjectives: "[...] affixes tara,
tama [...] are also added to the bases of interrogative pronouns,
but the meaning differs just a little from that of the respective bases (e.g. katara –
'which [of two]?'; katama – 'which [of many]
').
"
(f) Kena (instr.), kasmā (abl.) and kissa (gen.) are also employed adverbially,
meaning
"why?", "wherefore?"; kiṁ (acc.) is oftentimes used with the instrumental to express the meaning of
"what is the use of."
(g) Interrogative pronouns in all oblique cases may be used adverbially as well (Oberlies,
2019: 260, f.n.
1, 270; Palistudies, 2018f).

Indefinite Pronouns

Formation and Usage.(a) Indefinite pronouns don't refer to any person,
thing or amount specifically.
They are inexplicit, "not definite."
(b) Sometimes, substantive nouns are constructed from indefinite pronouns (e.g. kiñcanaṃ –
"defilement"; Oberlies, 2019: 283).
(c) They are formed in the following ways (Ānandamaitreya,
1993/2012: 81; Duroiselle, 1906/ 1997: 75; Perniola,
1997: 56–7; Warder, 1963/2001: 85–6):

• Addition of ci (cid before a vowel),
  cana (canaṃ is also found), api or pi to the interrogative pronouns (e.g. kiñci,
  kācana, kampi).
• Twofold repetition of the demonstrative or relative pronoun (e.g. so so –
  "anyone"; taṃ taṃ, in the sense of
  "several," "various").
• Joining a relative with an indefinite (e.g. yaṃ kiñci
  "whatever").
• Joining a negative with an indefinite (e.g. na kiñci
  "nothing").

Possessive Pronouns

Formation and Usage. (a) Some possessive pronouns form
from the base of the first and second personal pronouns by means of affixes īya and aka,
with occasional lengthening of the base vowel (e.g. mad + īya → madīya; mam + aka → māmaka –
"mine"; Duroiselle, 1906/1997: 77).
(b) The word attano (dative/genitive) can be used as a possessive pronoun (in all three genders,
singular as well as plural), meaning
"himself," "oneself," "myself," "yourself,"
agreeing with the subject of the clause or sentence (e.g. attāpi attānaṃ na upavadati –
"and he does not blame himself," MN II: 53 [MN 65]).
(c) As a possessive adjective, it stands for:
"his own," "her own," "my own" etc. (d) Other possessive pronouns and adjectives include: sayaṃ and sāmaṃ (
"oneself," "self," "myself" etc.; e.g. sāmaṃ diṭṭhaṃ –
"seen by oneself"); saka and sa (
"one's own"; Perniola, 1997: 57; Warder,
1963/2001: 187).
(e) The genitive of the personal pronouns for both persons also articulates a possessive sense (e.g. mayhaṃ,
tuyhaṃ, tassa; Perniola: 57).

*Pronominal Derivatives (Adjectives,

Adverbs)*

Adjectives (Duroiselle, 1906/1997: 77–8).
(a) Certain affixes added to pronominal bases form a great number of adjectives: di (dī),
disa, dikkha, disaka, risa intimate likeness and resemblance,
with lengthening of the base vowel (e.g. ta + di → tādi –
"like him," "like that," "such").
(b) As mentioned above: "[...] affixes tara,
tama [...] are also added to the bases of interrogative pronouns,
but the meaning differs just a little from that of the respective bases (e.g. katara –
'which [of two]?'; katama – 'which [of many]
').
"

Adverbs (Duroiselle, 1906/1997: 78–9).
(a) Certain suffixes added to pronominal stems form also a great number of adverbs: dā,
dāni, rahi express time (e.g. karahi –
"when?"; etarahi – "now").
(b) Adverbs of place are formed by means of suffixes to,
tra, tha, dha, ha, haṃ, hiṃ (e.g. kattha –
"where?"; yatra – "where?") – preceding a short vowel,
the t of tha is doubled.
(c) Suffixes thā, va, vaṃ, thaṃ,
ti are used to form adverbs of manner (e.g. tathā –
"thus," "like that"; kathaṃ – "how?
").
(d) Another va suffix is used to intimate time and cause (stem final a is lengthened),
appended to pronominal stems of ta,
ya, ki – final t is dropped since no consonant is permitted to stand as word final.
If it precedes a vowel, it is retained in the form of d (e.g. yāvadeva –
"ever so much," "as far as," "but only
").
(e) The ablative singular suffix tā is added to the mentioned forms too (e.g. yāvatā –
"as far as," "because").⁸⁷ (f) Suffix di is found merely in yadi (
"if"), ti in kati ("how many") and yati (
"as many").

⁸⁷ More examples of case forms used adverbially are listed above in the section
"Demonstrative Pronouns," pt.
(e).

Action Nouns

Formation and Usage. (a) The use of action nouns in Pāḷi is frequent – they are formed with affixes a,
i, ana, anā, aka, taṃ, tā,
ti, tta, ⁸⁸ added either directly to the root or the base (Perniola,
1997: 381).
(b) Action nouns express the generic verbal idea without reference to time or person (e.g. dassana –
"seeing") and, in the manner of any other noun,
assume the case form according to the respective syntactical
function they perform within the sentence (e.g. te kāyassa bhedā
[instr.] paraṃ maraṇā sugatiṃ saggaṃ lokaṃ upapannā –
"With the breaking up of the body, after death,
they are reborn in a good destination,
in the heavenly world,"
DN I: 40 [DN 2]; Hendriksen, 1944: 87).
(c) They usually govern the genitive case (e.g. brāhmaṇo
pokkharasāti bhagavato [genitive] adhivāsanaṃ viditvā [...] kālaṃ ārocesi –
"The Brahmin Pokkharasāti, having found out [about] the Blessed One's consent [...],
announced the time,"
DN I: 51 [DN 3]), but dassanāya stands also with the accusative (e.g. sammāsambuddhaṃ dassanāya upasaṅkamissāma –
"We would approach to see the Perfectly Enlightened One,"
MN II: 118 [MN 81]; Perniola: 381).
(d) Action nouns may take objects,
are modified by adverbs and are oftentimes continued with verbs (Hendriksen: 87).

Agent Nouns

Formation. (a) The affixes forming agent nouns are: a,
ana, aka, āvi, dha, i, in,
ina [after √ji], ka, ma, ratthu (tar),
ta, tra, tuka [after √gamu],
uka, ū⁸⁹ – they are appended to roots or bases (Collins,
2006: 113; Perniola, 1997: 378).
(b) Letter a [of roots] → i [occasionally] / __ ratthu (tar),
but final ā and e of roots remain unchanged (Ānandamaitreya,
1993/2012: 66).
(c) An i is also inserted under other circumstances (e.g. √bhāsa → bhāsitar; Warder,
1963/2001: 209).
(d) The affix aka is used after action nouns in ana to form agent nouns (Perniola: 155).
(e) The affixes of the past active participle may be superadded to these affixes (e.g. sutavā –
"he who has

⁸⁸ See the chapter "Kita and Taddhita Affixes
" for more details.

⁸⁹ See chapters "Kita and Taddhita Affixes
" and "Uṇādi Affixes."

heard").
(f) Agent nouns are oftentimes found as the final member of dependent-determinative compounds90 (Perniola: 378).

Usage – as Adjectives and Substantive Nouns. (a) Agent
nouns are frequently encountered in Pāḷi (more so in the earlier strata of the language) and may be translated as
"one who does [this or that]" or rendered simply by means of the English suffixes -er or -or,
denoting someone or something who/which does the action described
by the verb; i.e. the agent (e.g. tathāgato [...] daṭṭhāraṃ na maññati –
"The Tathagata [...] does not conceive the doer,"
AN IV: 16 [AN 4.24]; Hendriksen, 1944: 81–2; Perniola,
1997: 378).
(b) They may act as common nouns to take an object in the accusative,
genitive or dative case (e.g. bahujanamanāpā tathārūpiṃ vācaṃ [acc.] bhāsitā –
"the one who utters such speech which is pleasing to many people
"; bhinnānaṃ [gen.] [...] sandhātā –
"a conciliator [...] of those who are divided,"
DN I: 2 [DN 1]; Ānandamaitreya, 1993/2012: 66; Oberlies,
2019: 251, f.n.
4; Warder, 1963/2001: 209; Wijesekera,
1936/1993: 186).
(c) Agent nouns are also placed in apposition to a noun or
pronoun (e.g. ahamasmi brahmā [...] issaro kattā nimmātā –
"I am Brahma [...], the lord, maker and creator,"
DN I: 9 [DN 1]; Hendriksen: 84; Perniola: 379).
(d) They are also capable of functioning as adjectives (Ānandamaitreya: 66).

Usage – as Verbs and Predicates. (a) Agent nouns in Pāḷi
may express the main action of a sentence (e.g. samaṇo gotamo,
ito sutvā na amutra akkhātā imesaṃ bhedāya –
"The ascetic Gotama is not one who relates there what he has heard here for the division of those,"
DN I: 2 [DN 1]).
(b) They are also capable of denoting the action of a subordinate
clause (e.g. ahaṃ tena samayena purohito brāhmaṇo ahosiṃ tassa yaññassa yājetā –
"At that time, I was the king's high priest,
who was the performer of [or 'who performed
'] the sacrifice,"
DN I: 68 [DN 5]; Warder, 1963/2001: 211).
(c) Oberlies (2019: 251, f.n.
4) notes that "[s]yntactically they take either the accusative [...] or the genitive
" (e.g. [genitive] bhavissanti dhammassa aññātāro –
"There will be those who understand the dhamma,"
MN I: 114 [MN 26]; Hendriksen, 1944: 82).
(d) On account of the ambiguity of its

⁹⁰ See chapter "Compounds (samāsā)
" for details.

temporal sense, the agent noun may be used to express the past,
present and future tenses, thus denoting that which takes place
at any time⁹¹ (e.g. [potentially future] ayampi kho,
mahānāma, puggalo agantā nirayaṃ –
"Now this person, Mahānāma, will not go to hell
"; Hendriksen: 81, 83–4).

Grammatical Case (vibhatti)

Kinds of Cases

• 1. Nominative (paṭhamā or paccattavacanaṃ).
• 1. Accusative (dutiyā or upayogavacanaṃ).
• 1. Instrumental → ablative of instrument (tatiyā or karaṇavacanaṃ).
• 1. Dative (catutthī or sampadānavacanaṃ).
• 1. Ablative of separation (pañcamī,
     avadhi or apādānaṃ).
• 1. Genitive or possessive (chaṭṭhī or sāmivacanaṃ).
• 1. Locative (sattamī, bhummavacanaṃ or ādhāro).
• 1. Vocative (ālapana or āmantaṇavacanaṃ).

General Characteristics⁹²

(a) Noun case suffixes93 are affixed to nominal stems to indicate grammatical case.
(b) The traditional Pāḷi grammars acknowledge seven cases in total,
excluding the vocative for the overall tally (cf.
Yindee, 2018: 58); however, as mentioned earlier,"
in modern grammars, we may find it included among the total count,
therewith amounting to eight cases
" (Collins, 2006: 18–9).
(c) For the sake of completeness,
the list of the most elementary morphological case suffixes
– contained within the classical grammars – will be given in the following (Kacc 55; Rūp: 32; Thitzana,
2016: 186–9; Yindee, 2018: 58).
This listing is not an exhaustive guide to all the possible forms as shown in Table 3 in the
"Tables" section but would prove to be essential if one wishes to navigate the explanations

⁹¹ There seem to be, however,
no usages with an unequivocal future sense.

⁹² For an in-depth study about the various cases and their contextual applications see,
for example, Wijesekera (1936/1993).

⁹³ See Table 3 in the "Tables
" section for a comprehensive listing.

contained within the indigenous grammar books – one would also
better understand some derivations tendered in this grammar (singular endings are to the left,
plural ones to the right):

– Nominative si (→ o) yo (→ ā)
– Vocative si (→ ∅) yo (→ ā)
– Accusative aṃ yo (→ e)
– Instrumental nā (→ ena) hi (→ ebhi)
– Dative/Gen.
sa (∅ → s) naṃ (→ ānaṃ)
                                       94
– Ablative smā (→ mhā, ā)
                         95 hi (→ ebhi)
– Locative smiṃ (→ mhi, e)
                          96 su (final a [of stem] → e /__su)
                                                   97

To reiterate, the Padarūpasiddhi (Rūp: 29) and Duroiselle (1906/1997:
24) mention that whole syllables are at times elided for the
sake of meter or to facilitate pronunciation (e.g. abhiññāya sacchikatvā → abhiññā sacchikatvā).

Usage of the Cases

1. Nominative

• Subject (kattā lit.
  "agent") of sentences or clauses, active or passive.
  This is the main use of this case (Wijesekera,
  1936/1993: 39).
• Subject qualifiers: adjectives (guṇanāmāni),
  predicates (kiriyāni) or a term in apposition (e.g. [predicate] [...] saṅgati phasso –
  "The meeting [...] is contact," MN I: 80 [MN 18]).
• Items in a ti clause.
• Viewpoint (in the sense of "as"; e.g. pāmojjabahulo,
  dukkhassantaṃ karissati –
  "one will, as one having much joy, put an end to suffering,
  Dhp: 26, v.
  376).
• Text titles (e.g. dīghanikāyo).

⁹⁴ Vowel a [of stem] → (V̄).

⁹⁵ Suffix may remain unchanged.

⁹⁶ Suffix may remain unchanged.

⁹⁷ Suffix may remain unchanged.

• Exclamations (of abstract nouns).
• Hanging nominative, introduces another phrase without grammatical connection (Kacc 281,
  285; Collins, 2006: 19–20).
• The nominative can also be used instead of the locative (e.g. evaṃ kilesamaladhova,
  vijjante amatantaḷe.
  na gavesati taṃ taḷākaṃ, na doso amatantaḷe98 –
  "Just so there exists the pool of the deathless for the cleansing of the stains.
  If you don't search out that pool, it is not the fault of the pool of the deathless,"
  Bv: 6; bhikkhu nisinne mātugāmo upanisinno [...] hoti –
  "While the bhikkhu is sitting, the woman has sat down closely,"
  Vin I: 157 [Ay 1]).

2. Accusative

• Direct object, incl.
  goal of motion (kammaṃ) the main function of this case (Kacc 280; Wijesekera,
  1936/1993: 58).
• Internal direct object (e.g. "He sang a song
  ").
• With abstract endings ttaṃ and tā as object of verbs of motion or acquisition for change of state.
• Double accusative (e.g. taṃ ahaṃ brūmi brāhmaṇaṃ
  "Him I call a Brahmin," MN II: 203 [MN 98]).
• Viewpoint (in the sense of "in terms of,"
  "as"; e.g. yo ca abhāsitaṃ alapitaṃ tathāgatena abhāsitaṃ alapitaṃ tathāgatenāti dīpeti –
  "he who explains that which has not been said and spoken by the
  Tathagata as what was not said and spoken by the Tathagata,"
  AN II: 7 [AN 2.24]).
• Various adverbial uses:

⁹⁸ The respective commentary explicitly identifies
kilesamaladhova as a nominative employed in the sense of a locative: kilesamaladhovanti kilesamalasdhane,
bhummatthe paccattavacanaṃ (Bv-a: 47).
I am indebted to Bryan Levman, who pointed out this passage to me.
Both occurrences might be explained on different grounds,
so much so that this usage has to be considered unattested (Oberlies,
personal communication, October 3, 2020).

• Time during which (e.g. te tattha [...] ciraṃ dīghaṃ addhānaṃ titthanti –
  "They stay there for a long stretch of time
  "; Kacc 298).
• Extent of space (e.g. yojanaṃ "for a league
  "; Kacc 298).
• Manner (e.g. sādhukaṃ manasikarohi
  "Apply your mind [i.e. 'pay attention
  '] thoroughly!
  ", DN III: 75 [DN 31]).
• Object of various prepositions and postpositions: pacchā,
  antarā, yathā, vinā, santike, anu, abhi,
  paṭi (Collins, 2006: 20–3; Duroiselle,
  1906/1997: 155–6).
• May be used in the sense of the genitive,
  ablative, ⁹⁹ instrumental and locative (e.g. [locative]
  so [...] pubbaṇhasamayaṃ nivāsetvā pattacīvaramādāya gāmaṃ vā nigamaṃ vā piṇḍāya pavisati –
  "He [...], having dressed in the morning time and having taken his robe and bowl,
  enters a village or town for alms,"
  MN II: 63 [MN 67]; Kacc 275, 279, 297,
  306–307).

3. Instrumental

• The instruments (means) or things with which an action is completed;
  the fundamental use of this case (Kacc 279; Wijesekera,
  1936/1993: 108).
• Logical subject of passive verbs (e.g. svākkhāto bhagavatā dhammo
  "Well taught is the dhamma by the Blessed One,"
  DN III: 100 [DN 33]).
• Cause or reason (Kacc 289).
• Accompaniment (saddhiṃ and saha are not absolutely necessary;
  e.g. […] atha kho bhagavā āyasmatā aṅgulimālena pacchāsamaṇena yena sāvatthi tena cārikaṃ pakkāmi –
  "and then the Blessed One went to Sāvatthi with Ā.
  Aṅgulimāla as his attendant monk," MN II: 150 [MN 86]; Kacc 286).
• Manner.
• Attendant circumstances (e.g. abhibhū bhikkhu [...] dissamānenapi kāyena dhammaṃ desesi –
  "The bhikkhu Abhibhū [...] taught the dhamma with his body being visible,"
  SN I: 97 [SN 6.14]).

⁹⁹ With such words as dūra (
"distant," "far" etc.).

• Motion to a definite place can be expressed with yena-tena [
  "where-there"] constructions (e.g. aññatarā devatā [...] yena bhagavā tenupasaṅkami –
  "A certain deva went up to where the Blessed One was,"
  Khp: 2).
• Place (e.g. bhagavā dakkhiṇena passena sīhaseyyaṃ kappesi
  "The Blessed One lay down on the right side,"
  DN II: 57 [DN 16]).
• Time (e.g. tena samayena [...]
  "at that time [...]," DN II: 38 [DN 16]).
• Comparison (e.g. na tena seyyo sadiso ca vijjati
  "There exist none better or equal to him,"
  DN III: 65 [DN 30]; with saha at times in the sense of equality:
  "as").
• Other adverbial uses.
• With kiṃ in the sense of "what is the use of [...]?
  ", "away with [...]!", "no more of [...]!
  "; with alaṃ: "enough of [...]!","
  there is no need of [...]!"
  (Kacc 279, 286, 288–289; Collins, 2006: 23–7; Duroiselle,
  1906/1997: 156–7).
• Applied also in the sense of the ablative¹⁰⁰ and locative (Kacc 275,
  290, 296).

4. Dative

• Purpose, benefit, result.
  It primarily "denotes the thing with reference to which an action proceeds
  " (Kacc 109, 276; Wijesekera, 1936/1993: 156).
• With attha also in the sense of
  "for the sake of."
• Direction (e.g. appo saggāya gacchati
  "Few go to heaven," Dhp: 12, v.
  174).
• Time (e.g. dukkhamupenti punappunaṃ cirāya
  "For a long time, again and again, they undergo suffering,"
  Dhp: 24, v.
  342).
• Used also instead of the accusative and locative (Collins,
  2006: 27–8; Duroiselle, 1906/1997: 154–5).
• The dative of nouns in āya is often used in the sense of
  the infinitive and may be applied in a future sense (e.g. pākāya –
  "in order to cook"; Kacc 653; Duroiselle: 109).

¹⁰⁰ With such words as dūra (
"distant," "far" etc.).

– The infinitive is at times fully interchangeable with the dative of purpose (cf.
Warder, 1963/2001: 134; Wijesekera: 181).

5. Ablative

• The point from which, cause, origin,
  motive etc. the primary significance of the ablative case and the very opposite of the dative (Wijesekera,
  1936/1993: 192).
• (a) Measurement of distance and time (e.g. ito so,
  bhikkhave, ekanavutikappe yaṃ vipassī bhagavā arahaṃ sammāsambuddho loke udapādi –
  "Bhikkhus, ninety-one aeons ago [lit.
  'from now'] Vipassī arose in the world,
  the Blessed One, Worthy One, Perfectly Enlightened One,"
  DN II: 1 [DN 14]).
  (b) Used also with such words as dūra (
  "distant," "far" etc.), antika ("near
  ") and others of related meaning (e.g. āsanne ito naḷakāragāmo –
  "The village Naḷakāra is near from here,"
  MN II: 210 [MN 99]).
• Used with words indicative of purity,
  freedom, release, dissociation and
  "a little" as well as with the word pubba (
  "former," "before"; e.g. lobhaniyehi dhammehi suddho asaṃsaṭṭho –
  "It is pure and dissociated from greed-causing phenomena
  ").
• Comparison and, closely related to that,
  viewpoint (in the sense of
  "in terms of," "as" [with ablatives ending in to]; e.g. sārañca sārato ñatvā [...],
  te sāraṃ adhigacchanti –
  "Having known the essential as the essential [...],
  they attain the essential,"
  Dhp: 1, v.
  12).
• Certain adverbial forms: tasmā or tato (
  "therefore," "thence"), yasmā or yato (
  "whence," "because" etc.).
• Abstention from, with such words as ārati (
  "abstinence").
• Used also in the sense of the instrumental,
  accusative, genitive and locative (Kacc 275; Collins,
  2006: 28–31; Duroiselle, 1906/1997: 157–8; Wijesekera: 194,
  218).

6. Genitive

• (a) The genitive case is not merely used with verbs and substantive nouns but also with adjectives and adverbs,
  although normally it is found to qualify another noun.
  "It does so by assigning it to a particular class or description,
  or by distinguishing it as a part of a whole.
  So, the fundamental notion expressed by it is to mark the belonging to or being part of.
  This possessive or partitive application admits of the almost
  universal rendering of the gen. in Pāli as in the older languages by the English of
  " (Wijesekera, 1936/1993: 225).
  (b) So, the main sense is that of the first and second points below,
  with the remaining ones to follow also occurring in the language:
• Possessive (e.g. tathāgatassa parinibbānaṃ
  "the final extinction of the Tathagata,"
  DN II: 45 [DN 16]).
• Partitive, in the sense of "from among,"
  "of these."
• Subjective (e.g. [...] pacchimakaṃ [...] tathāgatassa vesāliyā dassanaṃ bhavissati –
  "This will be the Tathagata's [...] last sight of Vesāli,"
  DN II: 52 [DN 16]).
• Objective (e.g. māvamaññetha puññassa
  "You ought not disregard merit," Dhp: 9,
  v.
  122).
• Time (e.g. na cirasseva anupādāya āsavehi cittaṃ vimuccī
  "Just after no long time [or 'before long
  '],
  the mind was liberated from the defilements by non-clinging,"
  DN II: 16 [DN 14]).
• Used also instead of the accusative,
  ablative, instrumental and locative (Kacc 277,
  301, 304, 308–309; Collins, 2006: 31–4; Duroiselle,
  1906/1997: 154).

7. Locative

The primary meanings expressed by the locative case are:
"place at which" (proximity, domain),"
in which" (permeation), "on which,"
"into which" and "from which" (e.g. jalesu khīraṃ tiṭṭhati –
"The milk is in the water"; Kacc 278; Wijesekera,
1936/1993: 265), but it may comprise also the following:

• Comparison.
• Partitive, in the sense of "from among,"
  "of these."
• "With regard to," "in respect of,"
  "in re" (e.g. ariyasāvako rūpasmiṃ nibbindati –
  "The noble disciple is disgusted with regard to corporality,"
  MN I: 96 [MN 22]).
• Adverbial sense of space and time (e.g. sacepi [...] taṃ bhagavantaṃ dasasu yojanesu [...] –
  "even if [...] the Blessed One would be within ten leagues,"
  MN II: 137 [MN 84]).
• Adverbial (generally).
• Extensively used instead of the genitive,
  instrumental, dative and ablative (Kacc 278,
  302, 304, 310–313; Collins, 2006: 34–7; Duroiselle,
  1906/1997: 158–9).

Vocative

– Addressing listener (Collins, 2006: 37; Duroiselle,
1906/1997: 159).

Locative, Genitive, Accusative and Nominative Absolute

(a) A noun, pronoun or phrase together with a participle
in agreement with it is called an absolute construction; it functions
syntactically as an independent adverbial (i.e. qualifying) clause or phrase to denote time,
manner or attendant circumstance (though not invariably) in relation to a main clause.
(b) Its agent differs from the main clause or is impersonal (i.e. has no explicit agent; Collins,
2006: 37; Palistudies, 2018e; Wijesekera,
1936/1993: 302).
(c) The presence of a participle separates this construction
– having a subject and a predicate of its own¹⁰¹ –
from the main clause and is thus considered freed or detached
(Lat.: absolutus) from the remainder of the sentence; however,
a logical tie always remains.
(d) Absolute constructions occur with the locative,
genitive, accusative and nominative cases (order of frequency),
either with a present or past passive participle but never with the future passive participle or others,
such as the past active participle