Laghu 1

The Study Guide to
Laghu Siddhanta Kaumudi
Part 1 – Sajna Prakaranam &

Sandhi Prakaranam
Editor : Medhā Michika, AVG, Anaikatti
E Book published by:
Arsha Avinash Foundation

104 Third Street, Tatabad, Coimbatore 641012, India
Phone: + 91 9487373635
E mail: [email protected]
www.arshaavinash.in
The Study Guide to
Pā ini-Sūtra
through Laghusiddhāntakaumudī
Volume 1
(
!"#$%&' & *+#$%&' (
Medhā Michika
AVG Anaikkatti, 2016
Copyright © 2016 by Medhā Michika
All rights reserved.
The contents of this work may not in any shape or form be reproduced
without permission of Medhā Michika.
All profit from the sales of this book goes towards the activities initiated
by Śrī Pūjya Svamī Dayānanda Sarasvatī.
Electronic version of this book is available at:

www.arshaavinash.in
Printed version of this book is available at:

Arsha Vidya Gurukulam, Coimbatore, TN, India
www.arshavidya.in
Swami Dayananda Ashram, Rishikesh, UK, India
www.dayananda.org
Arsha Vidya Gurukulam, Saylorsburg, PA, USA
www.arshavidya.org
CreateSpace
www.createspace.com (Search by “Medha Michika”)
Amazon of your country
www.amazon.com etc. (Search by “Medha Michika”)
,' (
- ./0&12"3 4'5 -
6
- ./0789 4'5 -
This book is the first volume in a series titled

“The Study Guide to Pā ini-Sūtra through Laghusiddhāntakaumudī”.
As a traditional entry book into Pā inian Sanskrit grammar,
the Laghusiddhāntakaumudī of Varadarāja is studied widely, especially in India.
The “Study Guide to Laghusiddhāntakaumudī” series of books makes
Laghusiddhāntakaumudī easily accessible to students and teachers alike who desire to
gain a good grasp of Pā inian Sanskrit grammar in a proper traditional manner.
In this series, each Pā inian sūtra is clearly explained in Laghusiddhāntakaumudī
sequence with a diagram, word by word translation,
and translation of the vItti (the short commentary on the sūtra).
This series of books has been developed as study material in 39-month Vedanta and
Sanskrit courses conducted under Sri Pujya Swami Dayananda Sarasvati
at Arsha Vidya Gurukulam, Anaikkatti, Tamil Nadu, India.
i
Table of Contents
Purpose of this book .................................................................................................................. 1
6
Q4R+ST U6 3' –( Four aspects of this book:................................................................................ 2
(
Introduction to Z"$%&' .................................................................................................................... 5
What is Z"$%&? ........................................................................................................................... 5
(
^"*&4/3Z"$%&' ............................................................................................................................. 6
(
'*6 4`3' .......................................................................................................................................... 6
^"*&*4-'*6 45 (500 BCE) .................................................................................................................. 7
$"e"34-'*6 45 (400 BCE) ............................................................................................................... 8
^Th*i5 (150 BCE) ........................................................................................................................ 8
Introduction to QU"j"3/ `k ^"l5 ......................................................................................................... 9
Pā ini’s QU"j"3/ `k ^"l5 .............................................................................................................. 9
`k #n1o"5 (Types of Sūtras) ........................................................................................................ 11
(
pT-#$%&' (
........................................................................................................................................... 13
QU"j"3/ Q*q$"% `k MAP ................................................................................................................... 25
(
^*%n"s"-#$%&' .................................................................................................................................... 28
A. Interpretations of 5th, 6th, and 7th case-ending words ................................................ 30
B. Interpretations of words in '"4"*q$%& (Tot*u*q and To"*o*u*q) ................................... 35
C. Deciding v"*44, ( when there are many v"*44s( ................................................................ 38
D. Deciding wo12, when there are many wo12s ................................................................... 43
E. Deciding where w0' should be placed ........................................................................... 47
(
F. Extra step when xu&y/{u&y are replaced by Q&............................................................... 50
6 and u|*} are enjoined as wo12 ................................................... 51
G. Interpretation when 0&
H. Interpretation when •€, o/• y, and ‚ 6T are enjoined as wo12 .......................................... 53
ƒ„0y and Qû"o ............................................................................................................................. 55
Difference between *4s1q, Qû"o, and *43' `k ........................................................................... 56
Introduction to i•*6 }"t$…'o6 /........................................................................................................ 57
Two ^%†%" (traditions) to study ^"*&4/3Z"$%& ..................................................................... 57
Layout of i•*6 }"t$…'o6 / .......................................................................................................... 64
Goals of studying i•*6 }"t$…'o6 / ............................................................................................ 65
ii
Taking notes following i•*6 }"t$…'o6 / .................................................................................. 65
Prayer of i•*6 }"t$…'o6 / ........................................................................................................... 66
(
Q‡ !"-#$%&' ................................................................................................................................... 68
Q‡ QS-( *+5 ...................................................................................................................................... 91
Qˆ *+5 1) 3&-( *+5 “ ‰
6 ^"Š5”
6 ................................................................................................. 92
2 types of negation (#*Ts1q5) .................................................................................................. 100
Qˆ *+5 2) Q3u"3"u- *+5 “‹%31” .............................................................................................. 107
6 *+5 “ƒ^1Œ5” ................................................................................................... 116
Qˆ *+5 3) 0&-
Qˆ *+5 3) a. ƒ%& %•( ^%5 “$| Ž*}y5” .............................................................................................. 122
(
Qˆ *+5 3) b. ^ku`y "* }' “‹% p‹” ............................................................................................... 126
Qˆ *+5 4) u|*}- *+5 “$| Ž•$•'”( ............................................................................................. 130
Qˆ *+5 4) a. ‘e1qek’ 6 “ƒ^ •*T” ................................................................................................ 133
Qˆ *+5 4) b. u"*Ty$s for ‘e1qek’ 6 “Q“…*‹&/” ........................................................................ 136
Qˆ *+5 4) c. ƒ^ 0"y”*T q"T… “#"•y*T” ...................................................................................... 140
Qˆ *+5 5) ^%–^- *+5 “#1—T1” .................................................................................................. 144
Qˆ *+5 5) ^%–^- *+5 a. 2$˜"*os 6 “2$+56 ” ........................................................................ 146
Qˆ *+5 5) ^%–^- *+5 b. ,'"™9š “*2u1*‹” ........................................................................... 149
Qˆ *+5 6) o/• y- *+5 “o•e"*%5” .................................................................................................. 153
y ^- *+5 “‹%1›u” ................................................................................................. 155
Qˆ *+5 7) ^ku–
Qˆ *+5 8) #$| *Tn"u- *+5 “09Qœ'”( ......................................................................................... 157
Summary of the section starting from 1.1.52 Qi9›•Š .......................................................... 162
Summary of QS ( *+5 ................................................................................................................. 188
Understanding how the Pā ini system works ...................................................................... 189
Summary of !" `k s studied in QS-( *+ section .................................................................... 191
Summary of *`^"o/ `k s studied in QS-( *+ section ................................................................. 191
Summary of ^*%n"s" `k s studied in QS-( *+ section ............................................................... 191
Q‡ ‹i-( *+5 ..................................................................................................................................... 192
‹i-( *+5 1) š•/U
6 6 - *+5 “%"'Ÿ21T”1 ....................................................................................... 192
•
‹i-( *+5 2) ô"t-—Ÿ•- *+5 “u"0/25” .................................................................................... 203
‹i-( *+5 3) Q44"*
6 $- *+5 “‘T %6 "*%5”.................................................................................... 204
iii
‹i-( *+5 4) Last section .......................................................................................................... 208
‹i-( *+5 5) Q4€"%-
6 *+5 “‹*% u¡1” .......................................................................................... 221
‹i-( *+5 6) w0'5 ..................................................................................................................... 232
(
‹i-( *+5 7) 2nd 7• •-#$%&'....................................................................................................... 243
Summary of ‹¢ *+5 ................................................................................................................... 262
Q‡ *u 0y- *+5 .................................................................................................................................. 264
Summary of *u 0y *+ ................................................................................................................ 284
Index................................................................................................................................................ 294
Abbreviations:
AK – QZ3$925 Avyaya Kosa, a dictionary of indeclinables – Sri V. Srivatsankacharya – Samskrit
Education Society
Bh1 – n •'/Z"¤" #‡' n"0 – n/' 41 2"¥/
SK – * }"t$…'o6 /
B – R"i'49%'"
iv
Introduction
Purpose of this book
This book is a study guide to Pā ini-Sūtra through i•*6 }"t$…'o6 /.

There are two main purposes in writing this book:
1. To provide a bridge for students to the traditional way of studying the subject
matter.
Even though the subject matter itself is not that abstruse, the way it is presented in
tradition necessitates a bridge for modern day students. This book provides that bridge.
My professional background in computer system engineering helped a lot in this.
2. To re-connect the study of i•*6 }"t$…'o6 / to QU"j"3/.
As said in its prayer verse, the purpose of studying i•*6 }"t$…'o6 / is to get
introduced to Pā ini’s Z"$%& system presented in QU"j"3/.
However, i•*6 }"t$…'o6 / is commonly being studied today merely by memorizing
of all its u|*§s, without any cross-reference to QU"j"3/. By such manner of study, which has
become a convention among the majority of Z"$%& students, Pā ini’s Z"$%& system cannot
be properly understood.
Consequently, there has been much unfounded criticism of i•*6 }"t$…'o6 /, because
by studying it in this manner it does not give its students the benefit of studying Pā ini’s
Z"$%& system.
Whereas, what is actually at fault is how i•*6 }"t$…'o6 / is being taught, not
i•*6 }"t$…'o6 / itself. i•*6 }"t$…'o6 / is in fact a very sophisticated, well calculated, and time-
tested book of #*¨3" (studying method) for QU"j"3/. Other texts such as, $"*2$", #‡'"u|*§,
etc., are reference books alone. They do not provide this #*¨3".
To summarize, the proper way of studying i•*6 }"t$…'o6 / is to take it as a #*¨3"
book to QU"j"3/, and keep referring to QU"j"3/ in each step of the study. This method of
studying i•*6 }"t$…'o6 / is presented in this book.
1
Study Guide to Pā ini-Sūtra through i•*6 }"t$…'o6 /
6
Q4R+ST 6
U3' –( Four aspects of this book:
In the tradition, the four aspects of the book have to be made clear at the beginning
of the book itself, so that the reader can determine whether he or she should proceed to
read the book further. The four aspects of this book are as follows:
1. Q*q$"%/ – Who should read this book:

Anybody who is interested in Sanskrit grammar. Since I use a lot of terminology
used in the tradition and also of my own, the reader has preferably finished studying
“Sanskrit Grammar for Vedanta Students” volumes 1 to 3.
2. *us35 – Subject matter of this book:

Z"$%&', ( Sanskrit grammar, in general, and in particular, ^"*&4/3-Z"$%&', ( the
Pā inian grammar system developed by sage ^"*&*4 through which the entire Sanskrit
grammar is described in `k form.
3. #39—4' –( Usefulness of studying this book:

By completing the study of this book, three kinds of benefit will be gained by the
student.
The first benefit is clear understanding of Sanskrit language.
The second benefit is logical, objective thinking, blessed by ^"*&*4’s systematic and
ingenious presentation of the grammar rules.
The third benefit is getting proper introduction to other traditional scriptures in
Sanskrit, such as Vedanta 2"¥, Ayurveda, Jyotish, etc., as said in the last verse composed
by the author of i•*6 }"t$…'o6 /.
2"¥"t%1 #*uU"4" R"i"4" S9^$"*%$" « $| T" u%o%"—14 i•*6 }"t$…'o6 / -
4. ¬+5 – Connection:
There are many connections between many elements. The study of Sanskrit
grammar requires a lot of effort, and questions can tend to arise such as “why should I
2
Introduction
study Sanskrit?”, “Am I wasting my time studying Sanskrit?”, “I came to study Vedanta,
not Sanskrit”, etc., Therefore, to maintain enthusiasm for the study of Sanskrit grammar,
the student must have clear understanding of all of these connections and keep them in
mind whilst studying.
The connections are between:

u1o"t2"¥"*& and -| T' –( the original scriptures of Vedanta are in Sanskrit. Therefore,
in order to study directly from the original text, knowledge of Sanskrit language is an
inevitable requirement. (Note that the scriptures state that the Vedas are always in
Sanskrit, in every $®).
It should be noted here that one does not necessarily require knowledge of Sanskrit
language to gain the vision of Vedanta; provided that the teacher is capable of
communicating the vision effectively without using Sanskrit, and the student has adequate
preparedness and commitment.
However, if one has the opportunity to study under a traditional teacher of Vedanta
who teaches from the original texts in Sanskrit, then the benefits of learning the Sanskrit
language cannot be overstated.
(
-| T' and ^"*&4/3-Z"$%&' –( Sanskrit grammar can be studied using any system.
However, if the student is interested in studying traditional scriptures such as u1o"t2"¥,
study of the grammar should also be through a traditional system because this helps one’s
mind tune into the x*ss’ minds. And among all of the traditional systems of Z"$%&, the
only survivor today is ^"*&4/3-Z"$%&'. (
(
^"*&4/3-Z"$%&' and i•*6 }"t$…'o6 / – As the author of i•*6 }"t$…'o6 / says in its
opening prayer, i•*6 }"t$…'o6 / is meant for introducing ^"*&4/3-Z"$%& to those who are
new to it. i•*6 }"t$…'o6 / offers a well planned #*¨3" (studying method) to QU"j"3/. When
taught properly, i•*6 }"t$…'o6 / has proven over the last centuries to be a #*¨3" text
fulfilling its purpose.
3
i•*6 }"t$…'o6 / and this book – Even though i•*6 }"t$…'o6 / is meant for a student
who is new to traditional scriptures, it is still very difficult to study when one is foreign or
new to this traditional study method. This book bridges the gap between i•*6 }"t$…'o6 /
and the student by presenting the subject matter in a manner the student can grasp easily.
This book and the student – When the Q*q$"%/, who is interested in the *us3, studies
this book, he or she will gain the three-fold #39—4s described above.
4
Introduction
Introduction to Z"$%&' (
What is Z"$%&?
6
Z"*¨3t1 Z°±t1 2²"5 Q441 p*T 2²!"4—4$ Z"$%&' «( ('‹"n"³')(
That by which words are derived is called Z"$%&', ( which produces the knowledge
of the word. Z"$%& is also called ô2"¥.
(
6 "4"*'o' Qµ"¤"4'
#3´ «( ('‹"n"³')(
Z"$%& systematically presents #390, the usage of the language by *2Us, the x*ss,
'*6 4s, $*us, etc., The language precedes the grammar. The grammar is the description of the
6
language, not the prescription. Z"$%& teaches Z°*§ (derivation) of Sanskrit words of *2Us.
It is not meant for ƒ°*§ (creation) of the words.
*2“" Z"$%& ¶¡9 *47´ ·…*Ts T‡" «
$®š1*T s¸¹"*4 u1oŠ"º'y4/*s&5 -
Z"$%& is one of the six u1o"¹, auxiliary disciplines to the study of Veda. The six u1o"¹s
are *2“", $®5, Z"$%&', ( *47´', ( ¶¡5, and ·9*Ts'. ( Among them, Z"$%& is said to be the most
important - “#q"4 S s» (€¹1s 6 Z"$%&'”.(
6
All the Vedas, all the u1o"¹s, ¼|*T, ^%"&, #$%&œ½s, etc., are in Sanskrit language. When
one is committed to understanding these scriptures, learning of grammar is the inevitable
key in that pursuit.
5
^"*&4/3Z"$%&' (
The Sanskrit word for Pā inian grammar is ^"*&4/3Z"$%&'. (

^"*&*44" #9´ ^"*&4/3 Z"$%&' «( That which was taught by ^"*&*4 is called ^"*&4/3.
4.3.101 T14 #9´' «( 4.2.114 u|}"•5«)
The ancient sage grammarian ^"*&*4 authored a system in which the rules of the
entire Sanskrit language were presented. The system is in the form of small books, which
require a teaching ^%†%" (lineage) to unfold the meaning of the books. Students and
teachers in this ^%†%" are also called ^"*&4/3"5.
There were nine such systems in the ancient time. Hanumanji was a scholar of all
the nine, which are:
¾Œ S"Œ $"2$| ¿ $…'"% 2"$»"34' «(
(
"%€T S"*^2i 2"$i ^"*&4/3$' -
Among these nine grammar systems, ^"*&4/3Z"$%& is the only survivor in use today.
6
'*4`3' (
The founders of ^"*&4/3Z"$%& are known to be those three sages, '*6 4`3.
u"À$"% u%7*S n"³$"% ^Th*i' «(
(
^"*&*4 `k $"% S #&T9›*¼ '*6 4`3' -
I remain saluting to the three sages, u%7*S, the author of u"*Ty$, ^Th*i, the author of
'‹"n"³, and ^"*&*4, the author of ^"*&*4 `k .
The respective status of being #'"& (means of knowledge) in Z"$%& among the three
(
sages is in chronological order. Thus, ^Th*i’s '‹"n"³' gets (
the strongest #'"&•' (status of
being a means of knowledge). This is conveyed by this famous sentence: ƒ§%9§%'46 /4"
#"'"Á3' «(
6
Introduction
6 (500 BCE)
^"*&*4-'*45
^"*&*45 is a sage grammarian, who received '"‹1Â% `k s from lord *2u, understood the
whole picture of Sanskrit grammar, and composed Z"$%&2"¥s to teach the whole
grammar in his own system. This system, in which meta-language (artificial language) is
employed, is called ^"*&4/3Z"$%&. This is depicted in the following śloka which we chant
before the study of Z"$%&.
314"“% '"Ä"3'*q0Å '‹1Â%"T «(
$| ¿ Z"$%& #9´ T¼ • ^"*&431 4'5 -
My salutations to Pā ini, by whom the whole Z"$%&, the Sanskrit grammar, was
taught after receiving '"‹1Â% `k s from lord *2u.
He was born in 2"i"T%6 œ', ( known as Lahore in present time.
His works, the constituents of ^"*&4/3Z"$%&, are the following:
QU$ q"T^6 "lš 0&^"lÆ‡ •u S «
(
6 4 *2“" ^"*&4/3" Q'/ ¨'"T -
*i¹"42"
1) QU$' –( QU"j"3/ `k ^"l5 – A book which consists of about 4000 sūtras describing the
grammatical regulations of the entire Sanskrit language. The book has 8 chapters
and each chapter has 4 quarters. E.g., 1.1.1 u|*}%"o•S «(
2) q"T^6 "l5 – A book of q"Ts6 . By his observation of the language, forms of q"T 6 and their
meanings are defined in 10 classes. There are about 2000 q"Ts6 . E.g., nk §"3"' «(
3) 0&^"l5 – A book to show all the members of about 250 groups (0&) referred to in
QU"j"3/. E.g., (in QU"j"3/) 1.1.27 u"yo/*4 uy4"'"*4 « (in 0&^"l5) u"y*o0&5 – uy *uÂ ƒn
ƒn3…
6 4' –( A book to tell some of the regulations regarding genders of nouns.
4) *i¹"42"
6
E.g., (in ^*i¹"*q$"%1 (
) •Èt5 •"—tš « (•È, ( •, Q^, ( QS-#e3 ending words are masculine.)
5) *2“" – A book to describe v"4s, #3És, etc., for correct pronunciation. E.g.,
Q$6‹*u —y4/3"4" $Ál5 «
7
6 (400 BCE)
$"e"34-'*45
u%7*S5, also known as $"e"345, wrote u"*Ty$ on Pā ini’s QU"j"3/. u"*Ty$ is a type of
commentary on an original work, which has the purpose of correcting or improving that
( told in the following śloka.
which the author thinks is deficient. The definition of u"*Ty$' is
6
ƒ´"4´Ê7´"4" *St" 3` #uTyT«1
T œ½ u"*Ty$ #"ºu"y*Ty$!" *uS“&"5-
6 what is not
u"*Ty$ is a book in which consideration is done for ƒ´, what is said, Q4´,
said, and Ê7´, what is said in a difficult manner for understanding1.
The original work of u"*Ty$ in its entirety did not servive to today. The only u"*Ty$
known today are those u"*Ty$ which are quoted in '‹"n"³ by ^Th*i.
^Th*i5 (150 BCE)
Inferring from the dates, he may be a different person than the author of 390 `k or
the one known in w3u6 oË . He authored '‹"n"³', ( a commentary on u"*Ty$. His book is titled
so because of its volume and depth.
`k "‡Ì uÁ3yT 1 3` u"À5• `k "4 6 "*%*n5 «
€ô"*4 S uÁ3yt 1 n"³ n"³*uo9 *uÊ5 -
In which, meanings of the `k are explained by sentences following the word order
in the `k . Words used by n"³$"% are also explained. Such work is called n"³.
1
( - `k $"%1& 3o ( ƒ´, TŠ *us31 *St4' u"*Ty
ƒ´' p*T ( $•5 « Q4´' ( - “^Í'/ n314” `k $"%1& 3o ( Q4´'
6 p*T (
6 “n/*T5 n/5
6
Q*^ 39—4/3"” p*T u"Ty$"%1& ƒÎT1 « Q4´Š ( - ÊU" ƒ*´5 4 « Qu0t 6 Ê0y'–^1& ƒ´', (
*St4 u"*Ty$•5 « Ê7´' p*T
Qu0'4 1 *¡ÏqT" Š"T, ( 3T ( `k $"%1& $| UT3" ƒ´', ( TŠ ÐU/$%& *¨3T1 «ÐU/$%&"‡Ñ u"*Ty$ #uTyT 1 «
8
Introduction
k ^"l5
Introduction to QU"j"3/ `
Pā ini’s QU"j"3/ `k ^"l5
QU"j"3/ `k ^"l5 was composed by ^"*&*4-'*6 45. The subject matter of QU"j"3/ `k ^"l5 is
Z"$%&. “QU"j"3/” means “that which contains eight (QU4)( chapters (Qj"3)”. “ `k -^"l”
means “a book of sūtras”.
QU"j"3/ `k ^"l5 contains about 4000 sūtras. Each of the eight chapters (Qj"3) has
four quarters (^"os). Each quarter contains different numbers of sūtras. For example, the
sūtra “6.1.88 u|*}%1*S«“ is found in the 6th Qj"3, 1st ^"o, and it is the 88th sūtra in that ^"o.
What is `k ?
Six criteria are to be fulfilled for a literary work to be called sūtra, which are told in
the following śloka:
6 ' «(
Q®"“%' *¡Ïq "%u*ÒÂT9'Ó
QÆ9n'4u± S `k `k *uo9 *uÊ5-
“Those who know sūtras ( `k *uo5) know (*uÊ5) sūtra ( `k ')( to be: 1. Q®-Q“%', ( 2.
Q *¡Ïq', ( 3. "%uT, ( 4. *uÂT9'Ó
6 ', ( 5. QÆ9n', ( and 6. Q4u±'”(
(
1. Q®-Q“%' (that which uses the fewest number of syllables)
(
Q®"*4 (small, few) Q“%"*& (syllables) 3Š TT Q®"“% `k ' «( RºÔ/*‹ '" 5 (116B)
Since the entire 4000 sūtras are designed to be committed to memory, a sūtra
should use the minimum number of syllables possible. There is a saying “Qqy'"`"i"•u14
6
^`9„u 'Õ"t1 u•3"$%&"5« (by shortening the length by half '"`", grammarians consider it as a
celebration like forthe birth of an awaited son.)“
To achieve this Q®"“%•, one of the techniques Pā ini employs is “Q4u|6 *§5”,
repetition of a word from previous to subsequent sūtras for the sake of the intended
interpretation.
9
(
2. Q *¡Ïq' (that which is not ambiguous)
( *o‹ ( (to be doubtful, uncertain) + ´ = *+Ïq (that which is doubted)
'+
( Q *+Ïq' «( 4ØT°7s
4 *+Ïq' p*T 6 '" 5 (NT)
A sūtra should not be subject to ambiguity. Even if the number of syllables is
minimum, clarity should not be compromised.
(
3. "%uT (that which has significant meaning)
(
"% (essence, power) + 'T^• 6 (one who has ~) "%uT (
"%5 QŠ Q*Æ p*T "%uT ( `k ' (
Each sūtra is meant for making changes, such as giving a substitute, defining a term,
negating a rule, etc. Sūtra should have essence, or the ability to make a difference. In other
words, a sūtra should not be meaningless even when fulfilling other criteria.
4. *uÂT9'Ó' (
6 (that which has multiple applicability)
*uÂT5 (on all sides) 'Ó (
6 (facing) 3Š TT *uÂT9' 6
Ó `k ' «( RºÔ/*‹ '" 5 (116B)
Applicability of sūtra should be universal in many different occasions. To keep the
number of sūtras and the number of the syllables of the sūtra to a minimum, the scope of
sūtra should be maximum.
(
5. QÆ9n' (that which does not have exclamation or filler)
(
4 Æ9n' (interjection) p*T QÆ9n' «( 4ØT°7s
6 '" 5 (NT)
Sūtra should be free from interjections and exclamations.
(
6. Q4u±' (that which is not defective)
(
4 Qu±' (defective, imperfect) p*T Q4u±' «( 4ØT°7s
6 '" 5 (NT)
What sūtra conveys should be faultless. Sūtra should not be found defective or
imperfect.
Other literary works which conform to these criteria and are fit to be called `k ' (
( u1oZ" 5, 390 `k ' by
include ÚÛ `k ' by ( '‹*sy^Th*i5, etc.
10
Introduction
`k #n1o"5 (Types of Sūtras)
Sūtras in QU"j"3/ can be categorized into six types.

These six types of Pā ini Sūtras are told in the following verse:
!" S ^*%n"s" S *u*q*4 y3' ‘u S «
(
Q*To129›*q$"%š s¸*uq (
`k i“&' -
"There are six types of sūtras. They are the sūtras which give !" (definition), ^*%n"s"
(interpretation), *u*q (injunction), *43' (restriction), Q*To12 (extension), and Q*q$"% (topic)."
(
1. !"- `k ' (Sūtra which gives definition of a technical term)
A !", name, is given to a !/, the named. There is a certain relationship ( ¬+)
between !" and !/. This is called !"- *!- ¬+5.
In the metalanguage system of Pā inisūtras, !" is a technical term, an artificially
given name in order to achieve Q®"“%•, etc. In QU"j"3/, the !"- *!- ¬+ is established
by !"- `k . This is said: !"- *!- ¬+-QuR9q$ !" `k ' «(
Three things are to be known in understanding !"- `k .
1). !" (name)
2). !/ (named, meaning of the name)
3). !"-#o125 (usage of !")
1) !" (name)
Most of the !"s in sūtras are artificially given with the fewest number of letters,
such as •, *•, •,6 n, *», etc. These !"s themselves do not convey any meanings.
Some !"s were taken from other grammar systems already existed at Pā ini’s time
or before. In some exceptional cases, the meaning of the !" is the same as in the natural
language. They are called Qµ"‡ y- !", or '‹" !".
11
2) !/ (named, meaning of the name)

The literal meaning of !/ is: “that which has !" is !/”. What a !" indicates is a
!/. In a !" `k ', ( all the words other than !" are usually !/. For example, among all the
( 4"*
words in sūtra 1.3.2, i.e. “ƒô12 1 QS Q4 6 $5 pT”,( “pT”( is !" and the rest are !/.
3) !"-#o125 (usage of !")

After giving a name to a thing, the name should be used. This is conveyed by the
following u"À': ( “3" 3" !" " " ÝiuT/ « (whatever is a name, that has a purpose.)” !"s
(
which are defined in !" `k ' are utilized in other sūtras. Those sūtras are to be known as
!"-#o125.
For example:
1.1.60 Qo2y4 i9^5 «
1) !" i9^5
2) !/ Disappearance (Qo2y4')( of an existing thing
3) !"#o125 1.3.9 TŠ i9^5 «, 8.3.19 i9^5 2"$ÞŠ «, etc.
12
Introduction
(
pT-#$%&' (
Following Pujya Swamiji’s guidance, we shall commence our learning of Pā ini

grammar from the pT-( !" section. pT letters
( are defined by these seven !" `k s, plus one
*u*q `k .
6
1.3.2 ƒô12›1 —44"* $ pT «( ƒô12,1 pT 8(
1.3.3 ‹i•' «( 4
( «
1.3.4 4 *un´… T-6 -'"5
1.3.5 w*o*È y-»6-¸u5 « w*o5 8
1.3.6 s5 #e3Š « #e3Š 8
6 k«
1.3.7 S-»
1.3.8 i-2-ßuT*}T1 «
1.3.9 TŠ i9^5 «
The words along with number on the right indicate Q4u|6 *§s. These words flow
down until the sūtra with that number. If there is only number on the right, it means that
the whole sūtra flows down as Q4u|6 *§.
13
This is the first sūtra of pT-( !" section.
[ !" `k ']( 1.3.2 6 $ pT «(

ƒô12›1 —44"*
In the original texts of grammar, a nasalized vowel is pT. (
( Q44"*
ƒô12 1 7/1 QS 1/1 6 ( «
$5 1/1 pT 1/1
4 words in the `k , no Q4u|6 *§ is required.
• ƒô12 1 7/1 – ƒô12 means “the original texts of grammar“; details are shown below;
in Q*q$%&1 â'/.
• (
QS 1/1 – #e"‹"%5 QS; ( all the vowels; this is !/.
• 6
Q44"* $5 1/1 – A nasal sound; this is adjective to QS ;(
6
Q44"* $ is a !" defined by 1.1.8 'Ó 6
6 4"* $"uS49›44"* $5«.
• (
pT 1/1 – This is !".
[LSK] ƒô12 1 7/1 Q44"*

6 ( pT-( !$5 1/1 Š"T III/1
$5 1/1 QS 1/1 ( «
In the texts of grammar, a nasalized vowel is termed pT. (
The following śloka tells what are considered to be ƒô1252

#e3"5 *2u `k "*& ã"o12" w0'"Æ‡" «
q"T^6 "l9 0&1 ^"l ƒô12"5 #$/*TyT"5 -
1) #e3"5 – Suffixes taught in sūtras. E.g., ™* ,• *&*4•, 'T^• 6 (
2) *2u `k "*& – '"‹1Â% `k "*&. E.g., i• & «(
3) wo12"5 – Substitutes taught in sūtras. E.g., u • 6
4) w0'"5 – Augments taught in sūtras.
6 «, %' • 6 ¨/¸"3"' «(
5) q"T^6 "l5 – Those which are enumerated in q"T^6 "l5. E.g., Q • n*u
6) 0&^"l5 – Those which are enumerated in 0&^"l5. E.g., nuT • 6 «
2
To be more practical at the beginning, the above śloka is quoted in this book rather than the one
6 4' «( w0'#e3"o12" ƒô12"5 #$/*TyT"5 - 1. q"T^6 "l5, 2.
in äippa ī: [LSK *»å&/ æ] q"T 6 `k 0&9&"*ou"Î*i¹"42"
6 4', ( 6. w0'5, 7. #e35, and 8. wo125 are said to
('"‹1Â%) `k "*&, 3. 0&^"l5, 4. ƒ&"*o `k ^"l5, 5. u"Î*i¹"42"
be ƒô12.
14
Introduction
[ !" `k ']( 1.3.3 ‹i•' «( ~ ƒô12 1 pT (

( the teaching of Z"$%&'. (
The last consonant is termed pT in
‹i 1/1 ( « ~ ƒô12 1 7/1 pT 1/1

( Q•' 1/1 (
2 words in the `k ; 2 words as Q4u|6 *§

• ‹i ( 1/1 – #e"‹"%5 ‹i;( all consonants; this is !/.
• (
Q•' 1/1 (
– Qt1 nu' Q•' «( that which is at the end is Q•; this is adjective to ‹i;(
• 6
ƒô12 1 7/1 – From 1.3.2 ƒô12›1 —44"* ( ƒô12 means “the original texts of
$ pT «.
grammar“; in Q*q$%&1 â'/.
• (
pT 1/1 6
– From 1.3.2 ƒô12›1 —44"* ( This is !".
$ pT «.
( ‹i 1/1
[LSK] ƒô12 1 7/1 Q•' 1/1 ( Š"T III/1
( pT 1/1 ( «
In the teaching, the last consonant is pT. (
ƒo"‹%&"*4 (examples) are:

1) #e3"5 – ç» (, ^6 , ( *T^, ( T|S, ( 'T^• 6 (
2) *2u `k "*& – Q p ƒ & «( x { $ ( «
3) wo12"5 – Q4•™,( Þ^ (
6 ,( 4'6 (
4) w0'"5 – 3" »6 (, T$
(
5) q"T^6 "l5 – ¸6$|È $%&1 ( 1«
« o"& o"4
6) 0&^"l5 – w™ ( « 4È «(
15
(
The next sūtra gives exceptions to 1.3.3 ‹i•' «.
[*4s1q `k ']( 1.3.4 4 *un´… T¼ ( ( Q•' (

• 6 "5 « ~ ƒô12 1 pT ‹i
When the last consonant is Tu0y, , ( ', ( in ^6 or
( *T™,( it is not pT. (
( 1/3 « ~ ‹i 1/1
4 0 *un´… 7/1 T-• 6 -'"5 ( ƒô12 1 7/1 pT 1/1
( Q•' 1/1 (

• 4 0 – not
• (
*un´… 7/1 – *un*´ is !" for ^6 and *T™ ( 1.4.104 *un*´š «; in Q*q$%&1 â'/.
• ( 1/3 – T:• 6 S
T-• 6 -'"5 ( '5 S T¼
S ( for pronunciation. T • 6 is !" for
6 "5 (ID)«; the Q after ' is
Tu0y (T 6 S ( ' S). Total seven letters.

S
• (
‹i ( 1/1 – From 1.3.3 ‹i•' «.
• (
Q•' 1/1 (
– From 1.3.3 ‹i•' «.
• ƒô12 1 7/1 – ƒô12 means “the original texts of grammar“; in Q*q$%&1 â'/.
• (
[LSK] *un*´v"5 1/3 Tu0y- -'"5 1/3 4 0 pT5 1/3 «

In *un*´, even if they are consonants at the end, Tu0y, , ( and ' are
( not pTs.(
(
ƒo"‹%&"*4 (examples) are found only in ^6 and *T™,( because *un*´ is a !" given only
(
to ^6 and *T™:(
• In ^6 (
— ,( Q' ,( 2 ,( 8"' ,( *n ,( 8 ,( ™ ,( , ,( w' (
• In *T™ (
T ,( ‡ ,( u ,( ' ,( wT"', ( ‡" ,( w‡"' ,( è' (
16
Introduction
[ !" `k ']( 1.3.5 (

w*o*È y»6¸u5 « ~ ƒô12 1 pT q"T95
In the original texts of grammar, *È, »6, ¸6 at the beginning of q"T 6 are pTs.(
( q"T95 6/1
w*o5 1/1 *È-»6-¸u5 1/3 « ~ ƒô12 1 7/1 pT 1/1
2 words in the `k : 3 words as Q4u|6 *§
• q"T95 6/1 – From 1.3.1 nku"o39 q"Tu5« by 'Á¸k$‚ 6T0*T5 3, Q4u|6 *§ like the jump of a frog; the
1st case of the original is changed into 6th case by the force of the context; this is called
*un*´*u^*%&"' (modification of the case); in ¬+sé/ to w*o5.
• w*o5 1/1 – Beginning (of the q"T)6 ; adjective to *È-»6-¸u5.
• *È-»6-¸u5 1/3 – *È, »6, and ¸6; they are combinations of two letters each; »6 is not ƒ•*oT, (
abbreviation for »u0y. *È5 S »65 S ¸65 S *È»6¸u5 (ID); this is !/.
• (
[LSK] ƒô12 1 7/1 q"T95 6/1 w±"5 1/3 ‘T1 1/3 (*È-»6-¸u5 1/3) pT5 1/3 Š56 III/3«
In the grammar teaching, *È, »6, and ¸6 at the beginning of q"T 6 are termed pTs.(
ƒo"‹%&"*4 (examples) are found only in q"Ts6 :

• *È - *Èn/ n31 «
• —"ë uì*4•Ìs1 «
• »6 - »6,êk
(
• ¸6 - ¸6$|È $%&1 «
3
There are three types of Q4u|6 *§ :
6 y&' –( Just like the flow of a river, a word goes down from previous to subsequent sūtras.
1. Q4$s
This type is most commonly observed.
2. Q^$s y&', ( #*Ti9'"4$s
6 y&' –( Very rarely, a word can go upward to previous sūtras.
3. 'Á¸k$‚ 6T0*T5 – Sometimes a word jumps from one sūtra to another, passing over some sūtras
between. The literal meaning of the name is “going like the jump of a frog”.
17
[ !" `k ']( 1.3.6 (

s5 #e3Š « ~ ƒô12 1 pT w*o5
( the beginning of #e3 is pT. (
In the original texts of grammar, s at
(
s5 1/1 #e3Š 6/1 « ~ w*o5 1/1 ƒô12 1 7/1 pT 1/1
• s5 1/1 – The letter s; ( Q is for pronunciation (ƒí"%&"‡5y ); this is !/.
• #e3Š 6/1 – Suffix; in ¬+sé/ to w*o5.
• w*o5 1/1 – Beginning (of the #e3); adjective to s5.
• pT ( 1/1 – This is !".
( «
[LSK] #e3Š 6/1 w*o5 1/1 s5 1/1 p„!5 1/1 Š"T III/1
(
The s letter at the beginning of a suffix is pT. (
ƒo"‹%&"*4 (examples) are found only in #e3s:

• (
$| T-#e35
î4, ( s"$4 (
• T*}T#e35
³È, ( ï4 (
18
Introduction
[ !" `k ']( 1.3.7 (

S»• 6 k• « ~ ƒô12 1 pT w*o5 #e3Š
In the original texts of grammar, Su0y and »u0y at the beginning of #e3 is pT. (
(
S»• 6 k• 1/2 « ~ #e3Š 6/1 w*o5 1/1 ƒô12 1 7/1 pT 1/1
1 word in the `k ; 4 words as Q4u|6 *§
• ( (— ð
S»• 6 k• 1/2 – S:• 6 S »• 6: S S»• 6 k• (ID) « Su0y (S ¶ ( È( )( and »u0y (» ( l ( ¸ ( ñ ( &);( this is !/.

• w*o5 1/1 – Beginning (of the #e3); adjective to S»6 k .
• (
[LSK] #e3"o… 7/1 S»6 k 1/2 pT… 1/2 Æ5 III/2«

The letters of Su0y and »u0y at the beginning of a suffix are pTs.(
ƒo"‹%&"*4 (examples) are found only in #e3s:

• 4"*o-#e35
*&S ,( *&™ (
• (
$| T-#e35
S"4S ,( »$ ( , ¸, ¸6, *&*4•, Á3T ,( Áui
6 (
• T*}T#e35
( ¶, ¸T'S (
S*u,
• (
^6 -#e35
— ,( »"
• *T™-( #e35
&i (
• ¥/#e35
S"^ ,( »"^ ,( ¸"^ (
19
[ !" `k ']( 1.3.8 (

i2òT*}T1 « ~ ƒô12 1 pT w*o5 #e3Š
In the original texts of grammar, i,( 2, ( and $u0y at the beginning of #e3 which is not
T*}T is pT. (
(
i-2-$• 6 1/1 QT*}T1 7/1 « ~ #e3Š 6/1 w*o5 1/1 ƒô12 1 7/1 pT 1/1
• i-2-$• 6 1/1 – i5 S 25 S $• 6 : S i2$• 6 (SD) « i,( 2, ( and letters in $u0y; Q after i ( and 2 are
( for
pronunciation (ƒí"%&"‡5y ); this is !/.
• QT*}T1 7/1 – 4 T*}T5 QT*}T5 (NT) « that which is not T*}T; in Q*q$%&1 â'/.
• w*o5 1/1 – Beginning (of the #e3); adjective to S»6 k .
• ƒô12 1 7/1 – ƒô12 means “the original teaching of grammar“; in Q*q$%&1 â'/.
• pT ( 1/1 – This is !".
[LSK] T*}T-u—y-#e3"±"5 1/3 i2$u0"y5 1/3 pT5 1/3 Š56 III/3«

i,( 2, ( and $u0y letters at the beginning of a suffix other than T*}T are termed pT. (
ƒo"‹%&"*4 (examples) are found only in non-T*}T #e3s:

• 4"*o-#e35
ÀS ,( *ò^ ,( Às (
• (
$| T-#e35
Þ»6 ( , Þ^ ,( 2T|,• 2"4S ,( ´, ó", Ói ( , •È (
• (
^6 -#e35
2 ,( ™1 , ™* ,• ™ ,( *™
20
Introduction
After definitions of pT, ( this *u*q `k teaches that pT should

( be elided.
[*u*q `k ']( 1.3.9 TŠ i9^5 «

( elided.
That pT is
TŠ 6/1 i9^5 1/1 ।

• TŠ 6/1 – Pronoun To (, indicating the topic under discussion; in this context, the topic
is pT; ( v"431 90" sé/ by 1.1.49 sé/ v"431 90"«.
• i9^5 1/1 – This is $"3y, effect, of the sūtra; i9^ is defined as “disappearance of existing
letter” by 1.1.60 Qo2y4 i9^5«.
( «
[LSK] TŠ 6/1 pT5 6/1 i9^5 1/1 Š"T III/1
(
There is an elision of the pT letter.
( 2 is
For example, 2 of ( pT by
( 1.3.8 i2òT*}T1 « ~ ƒô12 1 pT w*o5
( #e3Š. And it is
elided by 1.3.9 TŠ i9^5«, resulting in “Q ”.(
21
2. Q*q$"% `k ' ( (Sūtra which tells the topic of following sūtras)
(
“Q*q$"%” means a topic, subject, title, government. Q*q$"% `k ' governs a certain
( understood as a common topic among the
number of sūtras which follow it. Q*q$"% `k ' is
following sūtras which it governs.
For example, the sūtra 6.1.72 *‹T"3"' «( in QU"j"3/ `k ^"l5 has the number 158 written
next to the sūtra. That means that the sūtras from 6.1.73 to 6.1.158 are under the topic of
*‹T". “ *‹T"3"'”( is a 7th case (*us3 â'/) singular form of *‹T" (phonetic connection).
The following sentence gives the definition of Q*q$"% `k ': (
(
€o12 1 Ýi2kÕ•1 *T ƒ§%9§%Ýi—4$•' Q*q$"% `k Š i“&' «(
In its own place (€o12 1 A7/1) when there is absence of usefulness (Ýi2kÕ•1 S7/1 *T S7/1)
( )
there is the status of giving the result in the following sūtras (ƒ§%9§%Ýi—4$•' 1/1
For example, the sūtra 6.1.72 *‹T"3"' «( does not produce any result, such as wo12
and #e3, by itself. However, this sūtra makes other sūtras meaningful when read together.
How can we know which sūtra is Q*q$"% `k ' ?( Pā ini gives a sūtra which defines
Q*q$"% `k '. (
1.3.11 €*%T14"*q$"%5«
€*%T14 3/1 (pônkTi“&1 T|T/3", the 3rd case interpreted as “in the form of …”) Q*q$"%5 1/1
(governing rule)
The word told in €*%T (one of the three accents of vowels) is Q*q$"% `k '. (
Important Q*q$"%- `k s
• 1.4.23 $"%$1 7/1 « up to 1.4.55
In this section the 6 $"%$s are defined.
• 3.1.1 #e35 1/1 « up to 5.4.160

In the 3rd, 4th , and 5th chapters, whatever is told in the 1st case gains #e3 !"
because of being '"4"*q$%& to this Q*q$"% `k .
• 3.1.2 ^%š 1/1, 0 « up to 5.4.160
22
Introduction
This `k covers the same scope as 3.1.1 #e35, indicating that the #e3 should come
after (^%5). This is why #e3 is called suffix. In these 3 chapters, the 5th case is the
*oÏ3901 ^Í'/ to tell what the #e3 should be suffixed to.
• 3.1.91 q"T95 5/1 « up to 3.4.117
From here (3.1.91) until the end of the 3rd chapter, #e3s are suffixed after q"Ts6 . Thus
(
all the $| T-#e3s and other suffixes in making *T™t are told in this section. Sūtras in
this section teach certain combinations of q"Ts6 and #e3s which are indicated in the
5th case and 1st case respectively.
• (
4.1.1 õ"^-#"*T^*o$"T ( « up to 5.4.160
5/1
(
In the 4th and 5th chapters, the #e3s are to be suffixed after #"*T^*o$s. All ^6 -#e3s,
¥/#e3s, and T*}T#e3s are taught under this section.
• 4.1.76 T*}T"5 1/3 « up to 5.4.160
• ( « up to 6.1.158
6.1.72 *‹T"3"' 7/1
“In the topic of phonetic connection.” Vowel sandhi rules are told in this section.
• 6.4.1 Q¹Š 6/1 « up to 7.4.97

Modifications on Q¹ are given in these 5 ^"os. After suffixing #e3, #$| *T (q"T,6 #"*T^*o$,
etc.) assumes the status of Q¹. Then the Q¹ can undergo some changes described by
the sūtras in this section. The modification on Q¹ is called “Q¹-$"3y”.
• 6.4 129 nŠ 6/1 « up to 6.4.175

This Q*q$"%- `k is within Q¹Š Q*q$"%. Q¹s can assume n- !" under certain
conditions. For that n- !$-Q¹, some modifications are described in this section.
• 8.2.1 ^ku`y "* }' 0,( 1/1 « up to 8.4.68

2nd, 3rd, and 4th ^"o of 8th chapter are called “*`^"o/”, a unit in which there are 3 ^"os.
(
This ^ku`y "* }' sūtra is told at the beginning of the *`^"o/, making two arrangements
regarding *`^"o/.
23
Arrangement 1: A sūtra in the *`^"o/ is not “seen” by preceding sūtras in *`^"o/. For
example, the sūtra 8.2.23 390"tŠ i9^5 « is not seen by 8.2.7 4i9^5 #"*T^*o$"tŠ «, as
in the case of the formation of the word “n0u"4”.(
The first 7 chapters and the 1st ^"o of the 8th chapter are together called “ ^"o âj"3/”,
a unit in which there are 7 chapters and one ^"o.
Arrangement 2: A sūtra in the *`^"o/ is not “seen” by another sūtra in ^"o âj"3/.
6 S «, as in
For example, the sūtra 8.2.7 4i9^5 #"*T^*o$"tŠ « is not seen by 7.3.102 *^
the case of the formation of the word “wö8"'”( .
These important Q*q$"% `k s are mapped in the following chart:
24
Introduction
QU"j"3/ Q*q$"% `k MAP
1st chapter 2nd chapter 3rd chapter 4th chapter 5th chapter 6th chapter 7th chapter 8th chapter
1st 2.1.1 '" 5 3.1.1 #e35 4.1.1 #"*T^*o$"T (

quarter 3.1.2 ^%š 4.1.2 €…— ...(
4.1.3 *¥3"' (
3.1.91 q"T95 4.1.76 T*}T"5 6.1.72 *‹T"3"' (
2nd 8.2.1 ^ku`y "* }' (

quarter
3rd 1.3.2 pT ( 2.3.1 *un*´5

quarter
4th 6.4.1 Q¹Š

quarter 1.4.23 $"%$1
3.4.78 *Tâ*÷ð... 6.4.129 nŠ
25
Study Guide to Pā ini Sūtra through i•*6 }"t$…'o6 /
(
3. *u*q- `k ' (Sūtra which gives an injunction to make a change in letters)
The word *u*q is comprised of *u + q" to ordain, enjoin. *u*q `k s give injunction,
which make some change in the process of –^* *} (grammatical formation of a word).
*u*q `k s are directly responsible for changing the form in –^* *}. Thus, *u*q `k s form
the main part of QU"j"3/, while other types of sūtras give support to *u*q `k s by providing
!", interpretation, etc.
The following sentence gives the definition of *u*q `k ': (
wo12"0'"*o*uq"3$ `k *u*q `k ' «(
The sūtra which ordains wo12, w0', etc., (wo12"0'"*o*uq"3$ `k 1/1) is called *u*q `k '. (
What *u*q `k s enjoin are:
1. wo125 (substitute)
2. w0'5 (augment)
3. #e35 (suffix)
Those things which are ordained are given in the 1st case in *u*q `k s. In the teaching,
they are together referred to as “$"3y”.
26
Introduction
(
4. *43' `k ' (Sūtra which restricts the application of another sūtra)
The following sentence gives the definition of *43' `k ': (

* }1 *T w%8'"&5 *43'"3 nu*T «
While something is already achieved (* }1 7/1 *T 7/1) by one sūtra, another sūtra is
still being started (w%8'"&5 1/1), or taught, in the same scope. Then that sūtra is (nu*T III/1)
for restricting (*43'"3 4/1), the application.
One of the characteristics of *43' `k is “‘u-$"%”, the word “‘u” in the sūtra or in the
u|*§. The word “‘u” means “only”. This word excludes all other possibilities.
For example, the sūtra 1.4.7 21s9 ø *Ó « gives all p-ending #"*T^*o$ the *•- !". Then
the next sūtra 1.4.8 ^*T5 '" ‘u « tells that ^*T gets *•- !" only (‘u) in '" . #"*T^*o$s like
./^*T and 4|^*T are '" s which end with ^*T word. They can get *•- !" by 1.4.7 21s9 ø *Ó «,
even without 1.4.8 ^*T5 '" ‘u «, since they are p-ending #"*T^*o$s. Still `k $"% (Pā ini)
started a new sūtra 1.4.8 ^*T5 '" ‘u « to exclude “^*T” which is not in a '" from getting
*•- !".
Another example is 8.2.23 390"tŠ i9^5 « and 8.2.24 %"T ( Š «. The first sūtra tells
“the last letter of a conjunct consonant-ending word should be elided.” The next sūtra tells
“after %,( only (
should be elided.” This implies that if any letters other than (
come after %,(
(
they are not elided. In other words, “%"T ^%Š Š ‘u i9^5, 4 QÕŠ «”. % (followed by (
becomes a conjunct consonant and the (

can be elided by the first sūtra 8.2.23 390"tŠ
i9^5 « without the second sūtra 8.2.24 %"T ( Š «. When that is the case, if the sūtra is not
understood as *43' `k , it will become redundant. Therefore the second sūtra should be
understood as *43' `k to exclude letters other than (
after % (from the elision. Because of
this sūtra, forms such as ù$y ( and ùú"y' from
( (
ù— y are possible.
27
5. ^*%n"s"- `k ' ( (Sūtra which tells how other sūtras should be interpreted)
The following sentence gives the definition of ^*%n"s" `k ': (

QZuv"3" Zuv" †"o$ `k ^*%n"s" `k ' «(
( ), a sūtra which brings a rule (Zuv"-
When there is no settled rule (QZuv"3"' 7/1
†"o$- `k ')( is ^*%n"s" `k '. (
For example, when there is more than one substitute enjoined in one place, that
situation is called “QZuv"”, the situation without settled regulations. In such a situation,
^*%n"s" `k s can provide a solution by giving interpretations. 1.1.50 v"4›1 t%T'5« is a
^*%n"s" `k which tells “pick the closest one in terms of v"4, etc.”. If that does not work,
another ^*%n"s" `k 1.3.10 3‡" û'4o16 25 '"4"' «( teaches “It should be in a respective order
when the numbers are the same.”
^*%n"s"-#$%&' (
Unlike * }"t$…'o6 /, i•*6 }"t$…'o6 / does not have ^*%n"s"#$%&, a section for teaching
main ^*%n"s" `k s. Since having a good grasp of some ^*%n"s" `k s is very much essential for
understanding all Pā ini-sūtras, it is ideal to have a collection of important ^*%n"s" `k s in
one place. In this guide book, the collection of important ^*%n"s" `k s are categorized and
explained. This section will be a very important reference as the study proceeds.
^*%n"s"- `k s can be categorized into these eight types by the function.
A. Interpretations of 5th, 6th, and 7th case-ending words
1.1.49 sé/ v"431 90" « 6th case indicates the v"4/ (v"431 90" sé/)
y «
1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ 7th case indicates what follows (^% â'/)
6
1.1.67 T¼"*oe§%Š « 5th case indicates what precedes (^ku^y Í'/)
B. Interpretations of words in '"4"*q$%& (Tot*u*q and To"*o*u*q)
1.1.72 314 *u*qÆotŠ « Supply “Qt” to *u21s&.

This operation is commonly known as Tot*u*q5.
28
Introduction
(
(u"þ) 3*¼4 *u*qÆo"o"u¢œ‹&1 « Supply “w*o” to *u21s&, ( when the words are in 7th case,
and one of them is indicating a letter.
This operation is commonly known as To"*o*u*q5.
C. Deciding v"*44, ( when there are many v"*44s(
1.1.52 Qi9›•Š « The last letter of what is presented in 6th case

1.1.53 *™í « Ditto, when the wo12 is *™T. (
1.1.54 wo15 ^%Š « The first letter of that which is following what is
presented in 5th case.
1 "i ( *2T ( uyŠ «
1.1.55 Q4$ All the letters of what is presented in 6th case,
1 "i ( or *2T. (
when wo12 is Q4$
D. Deciding wo12, when there are many wo12s
1.1.50 v"4›1 t%T'5 « The closest letter in terms of:

6 (#3É), 3. Q‡,y 4. #'"& ('"`")
1. v"4, or 2. 0&
1.3.10 3‡" ¤'4o16 25 '"4"' «( Respectively, in the order of appearance (¨'1&)
E. Deciding where w0' should be placed
1.1.46 w±t… »*$T… « (

*»T-w0' (
is placed at w*o, while *$T-w0' is placed at Qt.
( «
1.1.47 *'oS9›•"T ^%5 (
*'T-w0' is placed after the last vowel.
F. Extra step when xu&y/{u&y are replaced by Q& (
(
1.1.51 ƒ%& %^%5 « (
The Q& should ( .(
become %^%, the one followed by %/i
6 and u|*} are enjoined as wo12

6 }/ «
1.1.3 p$9 0&u| The word “p$5 6/1” should be added.
H. Interpretation when •€, o/• y, and ‚ 6T are enjoined as wo12
1.2.28 QSš« The word “QS5 6/1” should be added.
29
A. Interpretations of 5th, 6th, and 7th case-ending words
When a word in 5th, 6th, or 7th case in a sūtra is not understood by the meaning of *un*´
in the natural language, the following three sūtras give the special meaning of 5th, 6th, and
7th case used only in Pā ini-sūtras.
5 6 7
• 1.1.49 sé/ v"431 90" «

This sūtra brings the meaning “(the substitute should be done) in the place of ...” to
the word in 6th case. This is accomplished by bringing a word “v"4”1 in the u|*§.
• y «
1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
This ^*%n"s" `k gives a special 7th case meaning known as “^% â'/”. A word told in
7th case can be interpreted as “when … follows”. In u|*§, an extra word “^%1 7/1” can
be added to convey “when … is the one which follows,”. After adding the extra
word, the 7th case is understood as *T â'/.
• 6
1.1.67 T¼"*oe§%Š «
This ^*%n"s" `k is similar to the previous sūtra and known as “^ku^y Í'/”. The only
( 5th case” instead of 7th case and “ƒ§%Š” instead of “^kuŠ
differences are “T¼"T = y ”.
A word told in 5th case can be interpreted as “when … precedes”. In u|*§, “ƒ§%Š 6/1”
is added. ƒ§% and ^% are ^3"y3, synonyms. Hence “^%Š 6/1” can also be seen with this
^ku^y Í'/.
After adding the extra word, this 5th case is understood as *oÏ3901 ^Í'/ since “ƒ§%
(the following)” is a word requiring a referential point in 5th case.
30
Introduction
This ^*%n"s" `k gives an extra technical meaning to 6th case when the 6th case is not
understood in the senses of the natural language.
This kind of 6th case is called ^"*%n"*s$-sé/, or can be called v"431 90" sé/.
[^*%n"s" `k ']( 1.1.49 sé/ v"41390" «

The 6th case, whose ¬+ is not understood, should bring the word “v"4 1”, and
connect with the 6th case ending word.
sé/ 1/1 v"431 90" 1/1 «

• sé/ 1/1 – This is ƒ Ÿ1 3, subject.
• v"431 90" 1/1 – This is *uq13, predicate. “v"4”1 p*T ô14 3905 3Š"5 " v"431 90" (116B) « One
whose connection is with “v"4 1” word.
[K] p‹ 0 2"¥1 7/1 3" 1/1 sé/ 1/1 Q*43T390" 1/1 .k3T1 III/1 " 1/1 v"4 1390" 1/1 ‘u 0 nu*T III/1 «
In this teaching of Z"$%& of Pā ini, that 6th case is heard, and whose connection is
not determined, that 6th case is the one whose connection is to the word “v"4 1”.
“v"4”1 is in Q*q$%&1 â'/, meaning “in the place”. The v"431 90" sé/ is now ¬+1 sé/,
connected to “v"4 1”. Together the meaning is “in the place of …”
For example, in the sūtra 8.2.39 ði" —29›t1 « ~ ôŠ, it is understood that ôŠ 6/1 is
( , the word to which it is
connected to Qt1. Thus ^*%n"s" is not required. However for ði"' 6/3
connected is not told. In this case, this ^*%n"s" is required. The word “v"4 1 7/1” is supplied
and connected to the 6th case-ending word. The whole sentence is understood as follows:
( (v"4 1 7/1) —25 1/3 (wo12"5 1/3 Š56 III/3) «
ôŠ 6/1 Qt1 7/1 ði"' 6/3
(
In the place of ði ( at the end of ô, —2 should be the substitute.
Note that 1st case is understood as *u*q, such as wo12 and w0' without any ^*%n"s". A
(
verb such as Š"T III/1 and nu*T III/1 is added to the u|*§ to complete the sentence.
31
The next ^*%n"s" `k gives an extra technical meaning to 7th case when the 7th case is
not understood in the senses of the natural language.
This kind of 7th case is called ^"*%n"*s$- â'/, or can be called ^%- â'/.
[^*%n"s" `k ']( 1.1.66 y «

T*¼*ü*T *4*oýU 1 ^kuŠ
The $"3y, change, is enjoined in the place immediately before the 7th case-ending
word.
( p*T 0 *4*oýU 1 7/1 ^kuŠ
T*¼4 7/1 y 6/1 «
• T*¼4 ( 7/1 – By the force of p*T, it conveys “any word in 7th case”.
• p*T 0 – This word converts words in sūtra from 2²^% into Q‡^y %. Detail follows.
• ( say) + ´ (...ed) = that which is said without gap;
*4*oýU 1 7/1 – *4% ((without gap) + *o2 (to
in *T â'/, '"4"*q$%& to T*¼4. (
• y 6/1 – “$"3y'”( is supplied; in ¬+sé/.
^kuŠ
( $"3y' 1/1
[LSK] â'/*4oË241 3/1 *uq/3'"4' 1/1 ( u&"yt%1& 3/1 QZu*‹TŠ 6/1 ^kuŠ ( -
y 6/1 R9j' 1/1
An effect, which is being enjoined by presenting a 7th case-ending word, should be
understood as the one which is immediately previous to that 7th case-ending word.
( Q*S 7/1 « ~ *‹T"3"' 7/1

In the case of 6.1.77 p$5 6/1 3& 1/1 ( , with this ^*%n"s", it is
( the place of p$)( has to be known as immediately before QS”.(
understood that “$"3y (3& in
y ” is '"4"*q$%& to a word in v"431 90" sé/. The word told in 7th case should
The “^kuŠ
be in *oÏ3901 ^Í'/. In the example, QS5 5/1 ^kuŠ ( Š"T III/1
y 6/1 = p$5 6/1 (v"4 1 7/1) 3& 1/1 ( . For the
sake of simplicity, instead of “QS5 5/1 ^kuŠ

y 6/1”, “^%1 7/1” in *T â'/ is supplied, resulting in
( ^%, following”. The entire u|*§ will be:
“when QS is
( (wo125 1/1 nu*T III/1) Q*S 7/1 (^%1 7/1) *‹T"3"' 7/1
p$5 6/1 (v"4 1 7/1) 3& 1/1 ( «
( the substitute when QS is
In the topic of sandhi, in the place of p$,( 3& is ( following.
( is *us3 â'/. Thus ^*%n"s" is not required.
*‹T"3"' 7/1
32
Introduction
The role of “p*T” in Sanskrit – switching between Q‡ y^% and 2²^%
The indeclinable word p*T acts just as quotation marks do in English. The function
of p*T is to switch words or sentences between Q‡^y % (committed to the meaning) and 2²^%
(committed to the sounds).
In the natural language, words and sentences are Q‡^y % by default. Thus p*T converts
Q‡^y % to 2²^%. “0•” p*T 5 QuoT –( He said, “You may go”.
y «( tells that
In sūtra literature, words and sentences are 2²^% by default. 4.2.32 Q ñ1 $
ñ$ ( should be suffixed after Q* . Being 2²^%, Q* is understood as a string of letters: Q, 0, ( 4, (
p, not as the meaning of Q* , fire. Thus p*T converts 2²^% to Q‡^y %.
(
In the case of T*¼4 p*T, (
the p*T converts T*¼4 from “T, ( Q, , ( ', ( p, 4“( (2²^%) to “To (
with 7th case-ending”. To ( is a pronoun which can represent any word. Final meaning of
(
“T*¼4 p*T” is: “any word in 7th case”
6
The same applies for 1.1.67 T¼"*oe§%Š «, also.
33
This ^*%n"s" `k gives an extra technical meaning to 5th case by adding the word
“ƒ§%Š” connected to the 5th case-ending word.
There is no conventional name for this ^"*%n"*s$^Í'/. Yet, we shall call it ^ku^y Í'/ by
its context.
[^*%n"s" `k ']( 1.1.67 6

T¼"*oe§%Š « ~ *4*oýU"T (
The $"3y, change, is enjoined in the place immediately after the 5th case-ending word.
( p*T 0 ƒ§%Š 6/1 « ~ *4*oýU"T 5/1

T¼"T 5/1 (
3 words in the `k ; 1 word as Q4u|6 *§

• (
T¼"T 5/1 – By the force of p*T, it conveys “any word in 5th case”.
• p*T 0 – This word converts words in sūtra from 2²^% into Q‡^y %.
• ƒ§%Š 6/1 – In ¬+sé/. “$"3y'”( is supplied.
• (
*4*oýU"T 5/1 ( say) + ´ (...ed) = that which is said without
– *4% ((without gap) + *o2 (to
gap; after *un*´*u^*%&"', in *oÏ3901 ^Í'/, '"4"*q$%& to T¼"T; (
( $"3y' 1/1
[LSK] ^Í'/*4oË241 3/1 *uq/3'"4' 1/1 ( u&"yt%1& 3/1 QZu*‹TŠ 6/1 ^%Š 6/1 (= ƒ§%Š 6/1) !13' (
1/1
«
understood as the one which is immediately after that 5th case-ending word.
For example, in the sūtra 8.4.63 25 6/1 ¶5 1/1 Q*» 7/1 « ~ ð35 5/1 *‹T"3"', ( the 5th case-
ending word ð35 5/1 is not understood well. Even when the meaning is taken as *oÏ3901 ^Í'/,
( ? ƒT 0 ð35 5/1 ^kuŠ
questions such as “ð35 5/1 ^%Š 6/1 v"4 1 7/1 wo125 1/1 Š"T III/1 y 6/1?” may arise.
By this ^*%n"s", the word “ƒ§%Š 6/1 (^%Š 6/1)” is supplied in u|*§ and the whole sentence will
be understood as follows:
(
ð35 5/1 (^%Š 6/1) 25 6/1 (v"4 1 7/1) ¶5 1/1 Q*» 7/1 (^%1 7/1) *‹T"3"' 7/1
In the topic of sandhi, in the place of 2, ( which is after ð3, ( ¶ ( is the substitute when
Q» ( is following.
34
Introduction
B. Interpretations of words in '"4"*q$%& (Tot*u*q and To"*o*u*q)
When two words are in apposition ( '"4"*q$%&), and one word is adjective (*u21s&)
to another as substantive (*u21³), some interpretations may be required. The following two
^*%n"s" `k s give such interpretations.
• 1.1.72 314 *u*qÆotŠ «

This sūtra gives “Qt” as an Qj"‹"% (supplement word) between the two words in
apposition.
This sūtra is commonly known as “Tot*u*q”.
• (
(u"*Ty$')( 3*¼4 *u*qÆo"o"u¢œ‹&1 «
This vārtika gives “w*o” as an Qj"‹"% between the two words in apposition when
they are in 7th case and one of the words indicates letters.
This sūtra is commonly known as “To"*o*u*q”.
35
[^*%n"s" `k ']( 1.1.72 314 *u*qÆotŠ « ~ €Š –^Š

When *u*q is told with adjective, that adjective is represents for that which ends
with that adjective, and also for that adjective itself.
314 3/1 *u*q5 1/1 TotŠ 6/1 « ~ €Š 6/1 –^Š 6/1

• 314 3/1 – By which (adjective); it is understood as “314 *u21s&14”; in $%&1 T|T/3".
• *u*q5 1/1 – An injunction, rule, told by sūtra.
• ( *u21s4')( Qt1 3Š 5 Tot5 (176B) = 2² 'o6 "35 (group of letters), TŠ «;
TotŠ 6/1 – TT (=
“for the group of letters which ends with that adjective”; in ¬+sé/ connected to “ !"”
making a sentence “the adjective is a !" for the group of letters which ends with that
adjective”.
• €Š 6/1 – #"*T^*o$ € (its own) indicating *u21s&; in ¬+sé/, connected to “–^Š”
• –^Š 6/1 – Of the form; in ¬+sé/, also connected to “ !"”; “The adjective is also a
!" for its own form”
( TotŠ 6/1 !" 1/1 Š"T III/1

[SK] *u21s&' 1/1 ( €Š 6/1 S 0 7^Š 6/1 «
Adjective is a !" for a group of words ending with that adjective, and also for that
adjective itself.
For example, in the case of 7.3.101 QT5 6/1 o/•5y 1/1 3*È 7/1 « ~ Q¹Š 6/1, the two words
( taken as adjective to Q¹ without this ^*%n"s" `k ,
“QT5” and “Q¹Š” are in apposition. If QT is
the Q¹ should be only “Q”, which is not desirable. By this ^*%n"s" `k 1.1.72 314 *u*q5 TotŠ «,
( a !" for that which ends with QT, ( and also for QT itself”.
it is understood as “QT is ( In its
u|*§, the word “Qt” is supplied to the “QT”( to convey this meaning. Then they all together
are understood as “QotŠ 6/1 Q¹Š 6/1” or “Qot"¹Š 6/1”.
The formula is:
(
adjective: QT (short Q) + supplement: Qt (ending) + noun: Q¹
= Qot (short Q ending) + Q¹
= Qot"¹ (short Q ending Q¹)
36
Introduction
The next ^*%n"s" `k is Qû"o to the Tot*u*q. This operation is known as To"*o*u*q.
When two words in apposition are in ^% â'/, and the adjective is referring to letters
(
(Qi-œ‹&), the adjective is representative for a group of words beginning (w*o) with that
letters.
(u"*Ty$')( 3*¼*µ*qÆo"o"u¢œ‹&1 «
When Qi ( is mentioned as an adjective, and told in 7th case, then it should be
understood as “beginning with that letter”.
( *u*q5 1/1 To"o… 7/1 Qi-œ‹&1

3*¼4 7/1 ( 7/1
«
4 words in the u"*Ty$, other words are understood by the context.
• 3*¼4 ( 7/1 – “p*T â'/*4*oýU”1 is also understood; in *T â'/; '"4"*q$%& to Q¢œ‹&1;
• To"o… 7/1 – TT (= ( in
( Qi)( wo… 3Š 5 To"*o5 (176B) = 2² 'o6 "35 (group of letters), T*¼4 «;
^% â'/.
• (
Q¢œ‹&1 7/1 – Qi ( (any letter) œ‹&' (mentioned) (
Q¢œ‹&' (KT) T*¼4 «( in *T â'/;
( *u21s&1 7/1 Q¢œ‹&1 7/1 ^%1 7/1 *T 7/1 *u*q5 1/1 nu*T III/1, To" 0 “To"o… 7/1 *u21³ 1
[Medhā] 3o" 0 3*¼4 7/1
7/1
( «
^%1 7/1” p*T 0 u1*oTZ' 1/1
In which case *u*q is to take place when ^%*4*'§ is Qi ( (letter), and it is adjective to
another word, then it should be understood as “when the word beginning with that letter
is following”.
For example, in the case of 7.3.101 QT5 6/1 o/•5y 1/1 3*È 7/1 « ~ Q¹Š 6/1, “#e31 7/1” is
added as Q¹ presupposes #e3 following it. Now, these two words “3*È” and “#e31” are in
( an indicator of letters, Q¢œ‹&. Then by the ^*%n"s" (u"*Ty$')(
apposition in ^% â'/. And 3È is
3*¼*µ*qÆo"o"u¢œ‹&1 «, “w*o” word is brought in to the adjective “3*È 7/1”. Having put
them together, “3È"o… 7/1 #e31 7/1” (when 3È-beginning
( suffix follows) is understood.
The formula is:
( supplement: w*o (beginning) + noun: #e3
adjective: 3È +
= 3È"*o + #e3
37
C. Deciding v"*44, ( when there are many v"*44s(
When there is more than one letter in a word presented in 6th case, how to decide
which one letter should be replaced by the wo12 is told by the following four ^*%n"s" `k s.
6
क् ख् ग् घ्
Where should the wo12 be?

wo125
• 1.1.52 Qi9›•Š « ~ sé/

This sūtra tells “the wo12 should take place in the place of the final letter of what is
presented in 6th case.”
• 1.1.53 *™í« ~ sé/ Qi5 Q•Š

This sūtra tells “the last letter of what is presented in 6th case should be the v"4,
when the wo12 has ™ ( as pT.”( This sūtra is an Qû"o (exception) to 1.1.55 Q4$ (
1 "i*2T (
uyŠ«.
• 1.1.54 wo15 ^%Š« ~ sé/ Qi5

This sūtra tells “the first letter of what is presented in 6th case should be the v"4,
when a 5th case word is used in the sūtra.”
• (
1 "i*2T
1.1.55 Q4$ ( uyŠ« ~ sé/
This sūtra tells “all the letters of what is presented in 6th case should be the v"4,
( pT.”(
when wo12 is more than one letter, or has 2 as
38
Introduction
[^*%n"s" `k ']( 1.1.52 Qi9›•Š « ~ sé/

The last letter of what is presented by 6th case should be replaced by the substitute.
क् ख् ग् घ्
wo125
Qi5 6/1 Q•Š 6/1 « ~ sé/ 1/1

• Qi5 6/1 – #e"‹"%5 Qi ( any letter; in v"431 90" sé/ (1.1.49 sé/ v"431 90" «)
• Q•Š 6/1 – Qt1 nu5 Q•5« that which exists at the end is Q•5; adjective to Qi5.
(
[LSK] sé/*4*oýUŠ 6/1 Q•Š 6/1 Qi5 6/1 (v"4 1 7/1) wo125 1/1 Š"T III/1 «
The substitute should be in the place of the last letter of what is presented in 6th case.
For example, in the sūtra 8.2.23 390"tŠ 6/1 i9^5 1/1« ~ ôŠ 6/1, the substitute which
is elision (i9^5 1/1) is enjoined for ôŠ 6/1, which is in 6th case. 390"tŠ 6/1 is adjective to
ôŠ 6/1. The meaning of this sūtra is “There is elision (i9^5 1/1) of the word (ôŠ 6/1) which
ends with conjunct consonants ( 390"tŠ 6/1).” In this case, should the whole word be
elided? The ^*%n"s" `k 1.1.52 Qi9›•Š « brings a proper interpretation to achieve pU* *},
the desired form. According to this ^*%n"s" `k , only the last letter of what is in 6th case
should be the v"*44. (
39
[^*%n"s" `k ']( 1.1.53 *™í « ~ sé/ Qi5 Q•Š

The last letter of what is told by 6th case should be replaced by the substitute, when
(
the wo12 has ™ ( as pT letter. 1 "i ( *2T ( uyŠ «
This ^*%n"s" `k is an Qû"o for 1.1.55 Q4$
क् ख् ग् घ्
( 25
*™T-wo1
*™T ( 1/1 S 1/1 « ~ sé/ 1/1 Qi5 6/1 Q•Š 6/1

• (
*™T 1/1 (
– ™$"%5 pT 3Š (
5 *™T (116B) « This is an adjective to wo125. The wo12 which has
( called *™T. (
™ ( as pT is
• S 0 – This S brings down the entire sūtra: 1.1.52 Qi9›•Š «.
• Qi5 6/1 – #e"‹"%5 Qi;( any letter; in v"431 90" sé/ (1.1.49 sé/ v"431 90" «).
• Q•Š 6/1 – This is adjective to Qi5.
( Q4$
[LSK] *™T 1/1 1 "i 1/1 (
( Q*^ 0 Q•Š 6/1 ‘u 0 Š"T III/1 «
Even when there are more than one letter in wo12, if wo12 is *™T, ( it replaces only the
last letter of what is presented in 6th case.
When wo12 is Q4$ (

1 "i,( and at the same time the v"*44 should be only the last one
( work with this ^*%n"s" `k 1.1.53 *™í « as Qû"o to 1.1.55
letter, then the wo12 should be *™T to
1 "i ( *2T ( uyŠ «.
Q4$
For example, by the sūtra 6.1.123 Qu™ ( 1/1 ê9»"34Š« ~ 095 6/1 Q*S 7/1, the word 09
takes wo12 Qu™ ( when vowel follows. Being *™T, ( the wo12 Qu™ ( replaces only the last letter,
6 6 u"
,. Other examples are 6.4.77 Q*S q"T 6 9*%3™6 u™…«, 7.1.93 Q4™ ( …«, and so on.
40
Introduction
[^*%n"s" `k ']( 1.1.54 wo15 ^%Š « ~ sé/ Qi5

The first letter of what is following the 5th case-ending word should be replaced by
the substitute. This ^*%n"s" `k is an Qû"o for 1.1.52 Qi9›•Š «.
क् ख् ग् घ्
wo125
wo15 6/1 ^%Š 6/1 « ~ Qi5 6/1

• wo15 6/1 – #"*T^*o$ is w*o; adjective to Qi5.
• ^%Š 6/1 – In ¬+sé/ to wo15; “the beginning of what is following”.
• Qi5 6/1 – #"*T^*o$ is Qi,( any letter; in v"431 90" sé/.
( *u*‹T' 1/1
[LSK] ^%Š 6/1 3T 1/1 ( TT 1/1
( TŠ 6/1 wo15 6/1 R9j' 1/1
( «
That which is enjoined for what follows (after a 5th case-ending word) is to be
understood (as enjoined) for the beginning of that.
6
Because “^%Š” is a synonym to “ƒ§%Š” which comes from 1.1.67 T¼"*oe§%Š «, 5th
case is understood even though 5th case is not clearly mentioned in the sūtra or u|*§.
(
For example, the sūtra 6.3.97 *Ò-Qt%-ƒ^ ( « gives wo12
0Ë85 5/3 (^%Š 6/1) Q^5 6/1 T 1/1
(
in the place of Q^ when it comes after *Ò, Qt%,( or ƒ^ 0y. By this ^*%n"s" 1.1.53 wo15 ^%Š« ~
Qi5, the letter to be replaced is understood as the beginning letter of what is following.
( replaced.
Thus the Q of Q^ is
*Ò + Q^ (
*Ò + ^ ( (
6.3.97 *Ò-Qt%-ƒ^ 0Ë8›^ T «(
(
By 1.1.52 Qi9›•Š «, the substitute of at the last letter ^ was #"â.
By the Qû"o 1.1.54 wo15 ^%Š «, the beginning letter Q is substituted.
( Q
*Ò + ^ + 5.4.74 x %k 5k ^‡"'"4“1 « ~ '" "t"5
Ò/^ 6.1.101 Q$5 u&Ë o/•5y «
41
[^*%n"s" `k ']( 1.1.55 Q41$"i ( *2T ( uyŠ «

1 "i ( or *2T. (
The substitute should be in the place of the all letters when wo12 is Q4$
Q4$ (
1 "i ( 1/1 *2T 1/1 uyŠ 6/1 «
3 words in the `k ; no Q4u|6 *§ is required.
• 1 "i ( 1/1 – 4 ‘$5 Q4$
Q4$ 1 5 Qi ( 3Š 5 Q4$
1 5 (NT) « Q4$ 1 "i ( (116B) «
1 "i ( here.
This is an adjective to wo125. The wo12 which has many letters is called Q4$
• (
*2T 1/1 (
– 2$"%5 pT 3Š (
5 *2T (116B) «
( pT is
This is an adjective to wo125. The wo12 which has 2 as ( called *2T here.
(
• uyŠ 6/1 – in v"431 90" sé/ (1.1.49 sé/ v"431 90" «)
1 "i 1/1
[LSK *»å&/] Q4$ ( S 0 wo125 1/1 uyŠ 6/1 v"4 1 7/1 nuT5 III/2 «
( *2T 1/1
1 "i ( or *2T, ( should occur in the place of all the letters.

The substitute, which is Q4$
1 "i ( example:
Q4$
6 1 enjoins “nk” as an wo12 for “Q ”( q"T 6 when wqyq"T$
A sūtra 2.4.52 QÆ1n5k y « ~ wqyq"T$ 6
1 "i,( the entire group of letters “Q ”( will be replaced.

is intended. The “nk” being Q4$
1 "i,( if it is *™T, ( only
1.1.53 *™í « is an Qû"o of this sūtra. Even though the wo12 is Q4$
the last letter is substituted.
(
*2T example:
A sūtra 5.3.3 po' p2 «( enjoins “p2 1/1
( ” as an wo12 for “po'”( which is in 6th case. Of p2, (
( pT by
2 is ( 1.3.3 ‹i•' «.
( Even though the wo12 has only one letter (the number of letters
(
should always be counted without pT letters.), by being a *2T, ( the whole group of the
letters is replaced by this ^*%n"s" `k 1.1.55 Q4$ (
1 "i*2T ( uyŠ «.
Because this sūtra is ^% `k (that comes later) to 1.1.53 wo15 ^%Š «, even though the
1 "i ( or *2T, ( all the letters
place of replacement is told by 5th case-ending word, if wo12 is Q4$
( « ~ Q¹"T 5/1
are to be replaced. This is seen in the case of 7.1.9 QT5 5/1 (^%Š 6/1) *n 5 6/1 ¾ 1/1 ( .
( #"â as the letter to
The 5th case-ending word brings 1.1.53 wo15 ^%Š « and the first letter n is
( replaced with the help of
1 "i,( the entire *n is
be replaced. However, the wo12 being Q4$
(
1 "i*2T
1.1.55 Q4$ ( uyŠ «.
42
Introduction
D. Deciding wo12, when there are many wo12s
When there is more than one wo12 enjoined by a sūtra, how to decide which one of
the wo12s should be used is told by the following two ^*%n"s" `k s.
wo12 A, B, C, and D Which one of the wo12s to use?
• 1.1.50 v"4 1 Qt%T'5 «

This sūtra tells “when there is more than one possible choice, choose the closest
one.”
• ( o16 25 '"4"' «(
1.3.10 3‡" û' Q4
This sūtra tells “it should be in a respective order when the numbers of v"4/ and
wo12 are the same.”
43
[^*%n"s" `k ']( 1.1.50 v"41›t%T'5 «

(
When there are many possible wo12s, the most similar to the v"*44 should be
chosen.
v"4 1 7/1 Qt%T'5 1/1 «

2 words in `k , no Q4u|6 *§ is required.
• v"4 1 7/1 – v"4 here means # ¹, possibility of multiple wo12s; in *T â'/.
• (
Qt%T'5 1/1 – Qt% here means ”2, similar. To that, the superlative T*}T suffix T'^ is
added. The meaning is “the closest”. In which way it is the closest is discussed below.
(
[LSK] # ¹1 7/1 (= v"4 1 7/1) *T 7/1 ”2T'5 1/1 (= Qt%T'5 1/1) wo125 1/1 Š"T III/1 «
When there is possibility (of more than one wo12), the most similar one should be
the wo12.
By which aspect will the similarity be measured? There are four aspects for
measuring the similarity:
4 types of similarity (wt3y'):(
1. v"4T5 (by the point of articulation)

6
For example, in the case of “0¹" + p*T”, 6.1.87 wo ( 0&5« 6 is a name for
is applicable. 0&
three letters: Q, ‘, ,. Here, # ¹, multiple possibilities are found. The v"*44s( are w, whose
v"4 is $Ál, and p, whose v"4 is T" . The closest wo12 in terms of v"4 should be ‘, whose
v"4 is $ÁlT" , with the help of this ^*%n"s".
Other examples are 6.1.88 u|*}%1*S«, 6.1.101 Q$5 u&Ë o/•5y «, etc.
2. Q‡Ty 5 (by the meaning)

6 (
For example, in the case of “^%" S Q … *u±" S ^%*u±" (KT)«”, by 6.3.34 *¥3"5 ûo
6 6 ^%" becomes û6 "u, like masculine. Among all
™ ( '"4"*q$%&1 *¥3"'^k%&/*#3"*os«,
n"*sT^-"o4k
the words in masculine, a word which is the closest to the v"*44, ( namely ^%", in terms of
meaning should be the wo12, by this ^*%n"s". ^% is the most similar to ^%" in terms of
meaning.
44
Introduction
6
3. 0&T5 (by #3É, first examine w8t%, then R"ã)
( •È”( and “e— +
For example, in the case of “^S + ( •È”,( by 7.3.52 S—95 $6 *•ÁÁ3T95«, S (
(
and — are (
replaced by $u0y when *•T follows. # ¹, multiple possibilities, of wo12 are $,( Ó, ( 0, (
•, ( ™.( Here, v"4T5 does not work since every one has the same v"4. Then #3É is examined.
( Q®#"&, *uu"%, Â" , and Q•9s. The closest by #3É among $u0y is $.( In the same manner, — (
S is
is Q®#"&, u"%, 4"o, and •9s. The closest #3É among $u0y is 0. ( The results are ^"$ and e"0,
after ƒ^q" u|*}.
(
Other examples are 8.4.53 ði" — ð*2«, 8.4.62 ð39 ‹9›ÕT%Š"'«,( etc.
4. #'"&T5 (by '"`", the length of the sound)

(
For example, in the case of Qo¼ • becoming Q' 6 ,• and Qo"8"' becoming (
Q'k8"' by
(
8.2.80 Qo 9› o1 "yÊ o9 '5«, o ( becomes ' and the vowel after o ( becomes ƒu&y. By the aspect of
( •€, and o/• y ù is the
#'"&, the length, of this ^*%n"s", •€ ƒ is the wo12 when the v"*44 is
( o/• y.
wo12 when the v"*44 is
The v"4 should be examined first. If v"4 does not determine which wo12 to use,
any of the other three criteria can be utilized, without any priority amongst them.
45
[^*%n"s" `k ']( 1.3.10 ( '4o16 25 '"4"' «(

3‡" ™¤
When the number of what is enjoined and the number of the original are the same,
they are matched respectively (¨'1&) to the order.
v"4/ or !/ 1st 2nd 3rd 4th
wo125 or !" 1st 2nd 3rd 4th
( 3«
3‡" û' 0( Q4o16 25 1/1 '"4"' 6/
3 words in the `k ; no word as Q4u|6 *§
• 3‡" û' 0( – This is QZ3/n"u '" . *uœ‹u"À is “ û"' f/2/1
( (number) Q4*T¨Å (not
exceeding, not transgressing) p*T 3‡" û'”.( This word is used as adverb.
• Q4o16 25 1/1 – Any rules told in sūtras.
• (
'"4"' 6/3 – of two groups which have the same number of members
( ) *u*q5 1/1 (= Q4o16 25 1/1) 3‡" û' 0( Š"T III/1

[LSK] ' ¬+/ 1/1 (= '"4"' 6/3 ( «
The *u*q (Q4o16 2) which has connection between the same number should be
distributed according to the number.
' ¬+/ is adjective to *u*q5. ' ¬+/ means “the one who has the connection
between the same number.” '"4- û"4" ¬+5 ' ¬+5 « The connection ( ¬+) between
the same ( '"4) numbers ( û") is ' ¬+. ' ¬+5 QŠ Q*Æ p*T ' ¬+/ « ' ¬+ + p*4•
('•‡ y-T*}T-#e3, one who has …) = ''¬*+4, ( ''¬*+4 in
( masculine 1/1 is ' ¬+/.
For example, in the case of 6.1.78 ‘S5 6/1 Q3u"3"u5 1/3« ~ Q*S, the number of v"*44 (
and the number of wo12 are both four. And 1.1.50 v"4›1 t%T'5« does not work among them.
( ‘, ,, ¾, and ç are replaced by Q3, (
With the help of this ^*%n"s" 1.3.10 3‡" û'4o16 25 '"4"' «,
Qu, ( w3, ( and wu, ( respectively. When a word “¨'1&” is seen in a u|*§ of a sūtra, that is
because of this ^*%n"s".
Other examples are 1.2.27 ù$"i9› €o/•‚y 6T5«, 1.1.46 w±t… »*$T…«, 8.3.32 ™'9 •€"o*S
(
6 *4e'
™'& ( 8.4.40 Æ95 š4"
«, 6 š5«
6 etc.
46
Introduction
E. Deciding where w0' should be placed
(
Where w0' should be placed is decided By its pT-letter. Two ^*%n"s" `k s give rules
on this.
• 1.1.46 w±t… »*$T… «

(
This sūtra tells “*»T-w0' (
should be placed at the beginning, whereas *$T-w0'
should be placed at the end of that which is in 6th case.”
1 1
6 6
(
*»T-w0' (
*$T-w0'
(w*o-Qu3u) (Qt-Qu3u)
• ( «
1.1.47 *'oS9›Õ"T ^%5
This sūtra tells “When an w0' is *'T, ( it should be placed after the last vowel of
what is in 6th case.”
1 6
last *'T-(
QS ( w0'
Note that w0' itself does not have its own entity status. w0' always becomes part,
Qu3u of another entity, to which the w0' is enjoined.
.
47
[^*%n"s" `k ']( 1.1.46 w±t… »*$T… «

(
*»T w0' (
becomes w*o-Qu3u, while *$T w0' becomes Qt-Qu3u of the entity to
which the w0' is enjoined.
6 6
1 1
(
*»T-w0' (
*$T-w0'
w*o-Qt… 1/2 »-*$T… 1/2

• w*o-Qt… 1/2 w*o5 S Qt5 S w±t… (ID).
• »-*$T… 1/2 »5 S $ ( S »$… (ID), »$… pT… 3395 T… »*$T… (116B)«. Q after » ( is ƒí"%&"‡ y.
(
[LSK] *»T-*$T… 1/2
3Š 6/1 ƒ´… 1/2 TŠ 6/1 ¨'"T 0( w*o-Qt-Qu3u… 1/2 Æ5 III/2 «
(
For which entity *»T and (
*$T w0' are enjoined, they become the beginning part and
ending part, respectively, of that entity.
The word “¨'"T”( came from 1.3.10 3‡" ™¤ (
( '4o16 25 '"4"' «.
Example of *»T: ( 6.4.71 ™-( i™-( i|™ ( ¸6o"§5 « ~ Q¹Š

Q» (, which is *»T, ( is enjoined to Q¹ when ™ ( etc., are suffixed. The position of w0'
Q» ( is decided to be the beginning of Q¹ by this ^*%n"s" `k 1.1.46 w±t… »*$T… «, resulting in
QnkT, ( QnuT, ( Qn*u³T, ( etc.
Other examples are q»6 ( by 8.3.29 ¸5 * q»6 («, p» ( by 7.2.35 wqyq"T$
6 Š1¸ ( ui"o15«, w» ( by
7.3.112 wÁ4±"5«, 4»6 ( by 7.1.54 •€4±"^9 4»6 («, »6 ( by 7.1.52 w*' uy4"Ä5 »6 («, etc.
6 («
Example of *$T: ( 6.1.71 •€Š *^*T $| *T T$
6 ,( which is *$T, ( is enjoined to •€ vowel when *^T $|( T-#e3
T$ ( is following. The
6 ( is decided to be the end of •€ by this ^*%n"s" `k 1.1.46 w±t… »*$T… «,
position of w0' T$
6 , *u¼|e, *uÂ*—T, ( etc. Other examples are 8.3.28 &95 $6$-( »6$ ( 2*%«, etc.
resulting in #Æe
48
Introduction
[^*%n"s" `k ']( 1.1.47 ( «

*'oS9›•"T ^%5
(
*'T w0' is placed after the last vowel of the entity to which the w0' is enjoined.
6
last
QS (
1
(
*'T-w0'
( QS5 6/1 Q•"T 5/1

*'T 1/1 ( ^%5 1/1«

• (
*'T 1/1 ( 3Š
– ' pT ( (
5 *'T (116B)«
• QS5 6/1 – #e"‹"%5 QS, ( in *4q"y%&1 sé/; in ‘$uS4' to
( express —"*T, the class of QS; ( “among
all the vowels”.

• (
Q•"T 5/1 ( in *oÏ3901 ^Í'/ to ^%5.
– Qt1 nu5 Q•5, the last one, T¼"T «;
• (
^%5 1/1 – '"4"*q$%& to *'T-w0'. ( after the last vowel
Connecting all the words, “*'T is
among the all vowels.”
( 'j1 0 35 1/1 Q•5 1/1 T¼"T 5/1

[LSK] QS"' 6/3 ( ^%5 1/1 TŠ 6/1 ‘u 0 Qt-Qu3u5 1/1 *'T 1/1
( Š"T III/1
( «
(
*'T-w0' should be the end part after the vowel which is the last among all the
vowels.
Example of *'T: ( 7.1.58 p*oT9 4'6 q"T95«

(
4'6 , ( which is *'T, ( is enjoined to p*oT q"T

( .6 The position of w0' 4'6 is
( decided to be
(
after the last vowel by this ^*%n"s" `k 1.1.47 *'oS9›Õ"T ^%5«.
( .6 The 4'6 -w0'
For example, 4*o• is p*oT q"T ( whose content is “4”,( is placed after the
( Q + o (). Thus the augment “4”( comes after “Q”
last vowel among the content of the q"T 6 (4 +
( Q + 4+( o (”.
and the form will be “4¡ =( 4 +
49
F. Extra step when xu&y/{u&y are replaced by Q& (
(
By this ^*%n"s" `k , when xu&y and {uuy are replaced by the substitute Q& (Qu&y
, pu&y,
(
and ƒu&y) by a *u*q, that Q& should be followed by %1Ý or i.(
[^*%n"s" `k ']( 1.1.51 ƒ%& %•( ^%5 «

( the place of xu&y or {u&y should become %•^%, that which is followed by % (or i,(
Q& in
respectively.
( %•^%5 1/1 «
ƒ5 6/1 Q& 1/1
3 words in the sūtra, no Q4u|6 *§ is required.
• ƒ5 6/1 – #"*T^*o$ is x.; in v"431 90" sé/. “In the place of xu&y and {u&y”.
• (
Q& 1/1 – #e"‹"%5 with ^ku-y &$"%5, indicating Qu&y, pu&y, ƒu&y by 1.1.69 Q&*6 oT ( uy&Š
y
S"#e35«.
• (
%•^%5 1/1 – #e"‹"%5 %•, the first letter is % (of ‹3u%» (« and the end pT letter is Q• (by 1.3.2
6
ƒô12›1 —44"* $ pT«)( of i• &«.( %•: ^%5 3¼"T ( 5 %•^%5 (115B) from which ^% is %• (% (and i)( is %•^%5.
( «
[LSK] “x” p*T 0 *`2T5 6/1 !" 1/1 p*T 0 ƒ´' 1/1
It is said that x is a !" for 30 u&ys, 18 xs and 12 {s by 1.1.69 Q&*6 o„u&yŠ S"#e35«
(
and u"*Ty$ “x{u&y39*'y‡5 "uÁ3Ñ u"Î' «”.
[LSK] T (
"4 1 7/1 (x{u&yŠ 6/1 v"4 1 7/1) 35 1/1 Q& 1/1 ( ‘u 0 #uTyT 1 III/1 «
5 1/1 %•^%5 1/1 4 1/1
( to be there, it should be as %•^%5, that
In the place of xu&ys and {u&ys, when Q& is
which is followed by % (or i.(
For example, when Q of $| Ž is followed by x of x*}, they both are replaced by the
6 which is Q, by 6.1.87 w &5«.
most similar letter of 0&, 6 In this case x is being replaced by Q.
In such a case, the replacement Q should be %•^%, followed by %1Ý, by 1.1.50 v"4›1 t%T'5«.
Thus the final replacement is Q + % (because of 1.1.51 ƒ%& %•( ^%5«.
Similarly, when #*T^*o$ x is followed by ™* • (5/1 of ^6 ),( x, together with the
following Q, is replaced by ƒ by 6.1.111 xT ƒT«.( Assisted by 1.1.51 ƒ%& %•( ^%5«, wo12 will be ƒ
+ %,( resulting in ƒ% (+ . ( By 8.2.24 %"T ( Š« ~ 390"tŠ i9^5 and 8.3.15 Ó%u "439*uy —y4/35«, the
final result is ƒ5.
50
Introduction
6 and u|*} are enjoined as wo12

[^*%n"s" `k ']( 1.1.3 p$9 0&u| 6 u|*}5

6 }/ « ~ 0&5
The word “p$5 6/1” should be added when the place of 0&
6 and u|*} are not clear.
p$5 6/1 0&u|

6 }/ 1/2 « ~ 0&5
6 u|*}5
2 words in the `k , 2 words as Q4u|6 *§.

( 3` 0 0&u|
6 *}8"' 3/2
[SK] 0&u| ( ô' 1/1
6 }/ 1/2 *uq/31T 1 III/2 T` 0 “p$5 6/1“ p*T 0 s t' 1/1 ( ƒ^*TéT1 III/1 «
6 and “u|*}”, a 6th case-ending word

6 and u|*} are enjoined by the word “0&”
When 0&
“p$5” should be added.
[K] p$5 6/1 p*T 0 *$' 0( ? – wT-( ( '" 0 nkT III/1

“%-Zh4"4"' 6/3 ( «
Why “p$5” is required? – To prevent o/• y w, diphthongs (‘S),( and consonants being
6
subject to 0&/u|
*}.
6 395 7/2 « ~ Q¹Š 6/1 0&5

6 "qyq"T$
Example 1: 7.3.84 "uyq"T$ 6 1/1
6 7
Q¹ 6 /wqyq"T$
"uyq"T$ 6
1
6
0&
This sūtra gives rise to many ambiguities. Thus a few ^*%n"s" `k s are required to
6 1/1” as wo12 is told but the place of substitution is not
understand properly. First, “0&5
6 }/ « brings “p$5 6/1” into the interpretation.
clear. In this case, ^*%n"s" `k 1.1.3 p$9 0&u|
6 7
Q¹ p$ ( 6 /wqyq"T$
"uyq"T$ 6
1
6
0&
51
Now there are two 6th case ending words, “Q¹Š 6/1” and “p$5 6/1”. They need
another ^*%n"s" `k 1.1.72 314 *u*qÆotŠ «. The result is “p0tŠ 6/1 Q¹Š 6/1”.
Still, the v"4 is not clear. ^*%n"s" `k 1.1.52 Qi9›•Š « with 1.1.49 sé/ v"431 90" « is
required to decide the v"4 to be the last letter of the p0t-Q¹.
The final u|*§ made with the help of ^*%n"s" `k s will be: “p0tŠ 6/1 Q¹Š 6/1 Q•Š 6/1
Qi5 6/1 v"4 1 7/1 0&5 (
6 1/1 Š"T III/1 6 395 7/2 ^%395 7/2”.
6 "qyq"T$
"uyq"T$
In this example, “p$5 6/1” is used as an adjective to “Q¹Š 6/1”. (p$" Q¹5 *u21³T1«)
6 ^qŠ 6/1 S 0 « ~ Q¹Š 6/1 0&5

Example 2: 7.3.86 ^0ti•k 6 1/1 "uyq"T$ 6 395 7/2
6 "qyq"T$
6 7
Q¹ i• 6 6 /wqyq"T$
"uyq"T$ 6
1
6
0&
6 ^qŠ 6/1” is an adjective to “Q¹Š 6/1”. The ambiguity here is the v"*44. ( If
“^0ti•k
1.1.52 Qi9›•Š« is applied, even consonants can be the v"*44. ( When that is the case, by
6 will replace the consonant, but this is Q*4U, unwanted.
1.1.50 v"4›1 t%T'5«, the closest 0&
Here, “p$5 6/1” by this ^*%n"s" `k 1.1.3 p$9 0&u|
6 }/ « should be added. With “v"4 1 7/1” by 1.1.49
(
sé/ v"431 90" «, it is understood as “in the place of p$”.
6 7
Q¹ p$ ( 6 /wqyq"T$
"uyq"T$ 6
1
6
0&
6 ^qŠ 6/1 Q¹Š 6/1 0&5

The final u|*§ made with the help of ^*%n"s" `k s will be: “^0ti•k 6
1/1
(
Š"T III/1 p$5 6/1 v"4 1 7/1 ”.
This example is about Q¹14 p$ ( *u21³T1«.
52
Introduction
H. Interpretation when •€, o/•y, and ‚ 6T are enjoined as wo12
This sūtra is not covered in i•*6 }"t$…'o6 /.
[^*%n"s" `k ']( 1.2.28 QSš « ~ •€o/•y‚ 6T5

The word “QS5 6/1” should be added in u|*§ when wo12 is •€, o/• y or ‚ 6T, and the
place of replacement is not clear.
QS5 6/1 S 0 « ~ •€-o/•-y ‚ 6T5 1/1

2 words in the `k , 1 word as Q4u|6 *§.
( *uq/3T1 III/1 T` 0 “QS5 6/1” p*T 0 s t' 1/1
[SK] •€-o/•-y ‚ 6T-2²•5 3/3 3` 0 QS 1/1 ( ô' 1/1
( ƒ^*TéT1 III/1 «
Wherever a vowel is enjoined by the word “•€, o/• y or ‚ 6T”, in that place, the 6th
case-ending word “QS5 6/1” is supplied.
For example, in the case of 1.2.47 •€5 1/1 4^ 6 $1 7/1 #"*T^*o$Š 6/1 «, the word •€ 1/1
brings “QS5 6/1“ by 1.2.28 QSš« in u|*§. Then To* *q between QS5 6/1 and #"*T^*o$Š 6/1
(
should be brought by ^*%n"s" `k 1.1.72 314 *u*q5 TotŠ«, resulting in “for QS-ending
#"*T^*o$”. And the v"4 should be specified as the last letter by ^*%n"s" `k 1.1.52 Qi9›•Š «.
The final meaning of the sūtra will be: “•€ is the substitute in the place of the last letter of
(
an QS-ending #"*T^*o$ in neuter.”
Another example is 8.2.84 %"}kT 1 S « ~ u"ÀŠ »1 5 ‚ 6T5. In the case of ‹1 n0u4, ( the *» part
is Q4. ( By 1.1.52 Qi9›•Š«, the last letter could be the place for the substitute ‚ 6T, resulting
in {æ with 1.1.50 v"4›1 t%T'5«. To avoid this Q*4U# ¹, a 6th case ending word QS5 is added
by this ^*%n"s" 1.2.28 QSš«, to specify the v"*44. (
53
6. Q*To12 `k ' ( (Sūtra which extends the attributes of one thing to another)
Q*To12 means extension. Q*To12 `k extends the scope of a rule to areas where it was
not previously applicable.
For example, 1.1.56 v"*4uo"o129›4* q… « is an Q*To12 `k which gives the q'ys of v"*4
(what is going to be substituted) to an wo12 (substitute). To illustrate, “Þ^”( is an wo12 to a
(
$| T-#e3 “ó"”. By virtue of the sūtra 1.1.56 v"*4uo"o129›4* q… «, the q'ys of ó", such as the
(
status of being a $| T-#e3, the status of being a *$T, ( and its meaning “having done …”, etc.,
are extended to the wo12 Þ^. ( Thus Þ^ also
( enjoys all the same status and meaning.
(
Q*To12 `k s usually have uT-ending words, such as v"*4uT, ( ûT
6 , ( Qt"*ouT, ( T| T, ( *$ÒT, (
*™ÒT, ( etc., This T*}T#e3 “uT”( conveys the meaning “like ~”. Thus “~ uT”( means “like ~”.
(
The uT-ending word, being an adverb, is QZ3. In commentary, this is the typical style for
(
Q*To12 `k s: “wo125 v"*4uT nu*T «“ (the substitute becomes like the v"*4).
(
Nominalization of a uT-ending word (which is an adverb) is done by adding “n"u”
(the status of ~). The word will become “~ uo (n"u” (the status of being like ~). For example,
v"*4u "u, û6 "u, Qt"*ou "u, T| "u, etc., are the forms often seen in grammar books.
(
Q*To12 `k s can come without uT-ending (
( 6 »"*o89›*Ø&4 *™T
words as well. 1.2.1 0"™$ «(
and 1.2.5 Q 390"* » ( *$T «( are Q*To12 `k s which give *™Ò "u (the status of being like *™T)( and
*$Ò "u (the status of being like *$T)( respectively.
(
7. *4s1q `k ' (Sūtra which prohibits the application of another sūtra)
“*4s1q” means negation, or prohibition. It is also called #*Ts1q. A sūtra which negates
or prohibits the effect of other sūtras under certain conditions is *4s1q `k . This is said in the
following sentence:
^ku y `k $"3y*4s1q$ `k *4s1q `k ' «(
( 1.3.3 ‹i•' «gives
6 "5 « is a *4s1q `k to 1.3.3 ‹i•' «.
For example, 1.3.4 4 *un´… T¼ ( pT-(
6 "5 « prohibits giving the pT-( !" to the last
!" to the last consonants. 1.3.4 4 *un´… T¼
consonants when they are Tu0y, , ( or ' of
( *un*´.
54
Introduction
ƒ„0y and Qû"o
The concept of ƒ„0y and Qû"o is used throughout Pā ini sūtras to present rules
which cover the same scope of operation. ƒ„0y and Qû"o are relative terms. ƒ„0y is a
general rule, while Qû"o is a special rule whose scope is totally within the scope of ƒ„0y.
When such a relationship is observed between two rules, Qû"o takes effect by negating
ƒ„0y.
ƒ„0y5 Qû"o5
( "u$"25) (*4%u$"25)
Qû"o is called *4%u$"2 or Q4u$"2, that which does not have its own scope. On the
6
contrary, ƒ„0y is called "u$"2, that which has its own domain of operation. The Z°*§ of
ƒ„0y shows its nature: ƒ„|·T1 e·T1 R"q$- "u1 « That which is given up when there is
presence of its negator (R"q$ = Qû"o) is called ƒ„0y.
55
Difference between *4s1q, Qû"o, and *43' `k
*4s1q `k is told with reference to a *u*q `k . *4s1q `k prohibits the operation of the *u*q
by stating the word “4” and also certain things for which the *u*q is prohibited. For
example:
*u*q5 – All the residents of this gurukulam get an apple.
*4s1q5 – Teachers do not get.
Qû"o `k enjoins a different $"3y to certain things which are in the subset of the scope
of ƒ„0y `k . For example:
ƒ„0y*u*q5 – All the residents of this gurukulam get an apple.
Qû"o5 – Teachers get an orange.
What is told by *43' `k is already * }, accomplished by *u*q `k , thus it may look

redundant. (* }1 *T w%8'"4 *43'"3«) The purpose is to exclude any other things. For
example:
*u*q5 – All the residents of this gurukulam get an apple.
*43'5 – When it rains, students get an apple.
56
Introduction
Introduction to i•*6 }"t$…'o/

6
i•*6 }"t$…'o6 / is a small (i•)6 i.e. abridged version of * }"t$…'o6 /, which is a #*¨3"œ½
(a book providing a studying method) to ^"*&4/3Z"$%&.
Two ^%†%" (traditions) to study ^"*&4/3Z"$%&
1. #"S/4-^%†%"
In this tradition, the whole QU"j"3/ is memorized first, before studying the meaning.
It is referred to as Qé"j"3/-¨'-Q4 6 "%/, that which follows the order of Qé"j"3/.
$"*2$" u|*§5 is the main book to be studied in this ^%†%".
About $"*2$" u|*§5
$"*2$" u|*§5 was composed by —3"*oe5 and u"'45 in the 7th century. This book gives
u|*§5, a short commentary on each sūtra in the order of QU"j"3/ sūtras.
There are two main commentaries on $"*2$" u|*§5
1 R*6 }5 (8th century)
1. Õ" 5 by —4Œ
2. ô'h%/ by ‹%o§5 (12th century)
#‡'"u|*§5, the first (#‡') round (wu|*§) of grammar study, by ÚÛo§*—!" 56 is based on
$"*2$" u|*§5. This is used as a handy reference because it gives ô•1o5, Q4u|6 *§5, '" "‡5y , `k "‡5y ,
ƒo"‹%&', ( and Hindi translation for each sūtra.
2. 4Z^%†%"
It is referred to as #*¨3"-Q4 6 "%/, or –^* *}-Q4 6 "%/, that which follows method to
achieve certain forms.
The famous work under this type of study is * }"t$…'o6 /.
About * }"t$…'o6 /
* }"t$…'o6 / was written by n 9—/o/*“T5 in the 17th century.
57
This book follows #*¨3", the method, or discussion of achieving the final form. Here
sūtras are presented in the process of –^* *}, achieving the final form, and not in the order
of Qé"j"3/. The entire 4000 sūtras, 2000 q"Ts6 , and ƒ&"*o sūtras are presented in this manner.
Its commentary for each sūtra consists of two parts: u|*§ part and –^* *} part, which
will be understood when actually studied.
The book consists of different topics such as !"#$%&', ( *+#$%&', ( *T™t#$%&', ( etc.,
Under each topic there are selected examples which are unique by nature and are well-
calculated to show the features of the sūtras. It is easy to study and teach this book
because one can safely follow the book without going wrong.
There are two main commentaries on * }"t$…'o6 /:
1. R"i'49%'" by u" o16 uo/*“T5
1 %€T/
2. T R9*q4/ by !"4Œ
About i•*6 }"t$…'o/

6
u%o"%"—"S"3y wrote i•*6 }"t$…'o6 /, which consists about 1200 sūtras and vārtikas
together from * }"t$…'o6 /.
n •'/Z"¤" by n/' 41 2"¥/ is a very good commentary on i•*6 }"t$…'o6 /.
This Study Guide to Pā ini-Sūtra through i•*6 }"t$…'o6 / closely follows the
Gitapress version of i•*6 }"t$…'o6 /, which has become known as a standard book.
58
Introduction
Structure of i•*6 }"t$…'o/

6
Now let us see the index of i•*6 }"t$…'o6 /.
• ( the topics) Q4¨'5

Q‡ i•*6 }"t$…'o6 /-v(obtains in)-*us3"&"' (of 6 (index)
• *us3"5 (topics)
• ^|é(page)-Q "5 (number)
These topics are categorized as sections (#$%&s) as follows:

1. !"#$%&' (
- !"#$%&' (
!" there means “technical term” used only in this Pā ini grammar system. This
section lays out all the letters and their varieties by introducing related technical terms.
In this section, '"‹1Â% `k "*&, 14 `k s (13 !" `k s and 1 *u*q `k ) and one u"*Ty$ are taught.
Some important *2“" `k s are also taught.
All the `k s as well as u|*§s in this section are very important. It is strongly
recommended to commit this whole section to memory.
2. *+#$%&' ( Three types of *+ rules are taught in this section.

– QS ( *+5 (vowel sandhi)
æ – ‹i ( *+5 (consonant sandhi)
– *u 0y *+5 (visarga sandhi)
59
6
3. Rt#$%&' (
! – Q—t^*• 6 ¹"5 (vowel-ending masculine nouns)

" – Q—t¥/*i¹"5 (vowel-ending feminine nouns)
# – Q—t4^ 6 $* ¹"5 (vowel-ending neuter nouns)
$ – ‹it^*• 6 ¹"5 (consonant-ending masculine nouns)
% – ‹it¥/*i¹"5 (consonant-ending feminine nouns)
þ – ‹it4^ 6 $* ¹"5 (consonant-ending neuter nouns)
– QZ3"*4 (indeclinables)
Declension of nouns is taught in this section in 6 parts, known as s¸ (*i¹"5, which
consist of #"*T^*o$s ending with vowels and consonants in 3 genders.
6
Note that QZ3s are also Rts.
Up to this point is considered to be the “first half (^ku"yq y)” of i•*6 }"t$…'o6 /.
4. *T™t#$%&' (
Conjugation of verbs is taught by class (“conjugation”), from 1 to 10. Each q"T 6 is
conjugated in 10 i$"%s in $Ty*%#390 only. This section takes the maximum time to complete.
Even if the student’s time is restricted, the entire &"o35 (1st conjugation) and the first q"T 6 of
each of the remaining conjugations should be studied.
– &"o35 (1st conjugation, q"Ts6 starting with nk in q"T^6 "l5)
æ – Qo"o35 (2nd conjugation, q"Ts6 starting with Qo ( in q"T^6 "l5)
6
– —‹9e"o35 (3rd conjugation, q"Ts6 starting with º in q"T^6 "l5)
( q"T^6 "l5)
! – *ou"o35 (4th conjugation, q"Ts6 starting with *ou in
" – €"o35 (5th conjugation, q"Ts6 starting with 6 in q"T^6 "l5)
# – To6 "o35 (6th conjugation, q"Ts6 starting with To6 ( in q"T^6 "l5)
( q"T^6 "l5)
$ – 7q"o35 (7th conjugation, q"Ts6 starting with 7q in
( q"T^6 "l5)
% – T4"o35 (8th conjugation, q"Ts6 starting with T4 in
þ – *¨3"o35 (9th conjugation, q"Ts6 starting with ¨/ in q"T^6 "l5)
6
– S%"o35 (10th conjugation, q"Ts6 starting with S%6 (in q"T^6 "l5)
(
– Á3t#*¨3" (conjugations of *&S-#e3-ending q"Ts6 )
60
Introduction
(
æ – üt#*¨3" (conjugations of 4-#e3-ending q"Ts6 )
– 3™t#*¨3" (conjugations of 3™-( #e3-ending q"Ts6 )
! – 3™ ( '*¨3" (conjugations of 3™-( #e3-ending q"Ts6 after eliding the 3™)(
" – 4"'q"Tu5 (q"Ts6 derived from nouns)
# – $Á¸ (u"o35 (q"Ts6 in the category starting with $Á¸6)
$ – wö4^1 o#*¨3" (instances in which wö4 1ô-#e3s are used)
% – ^%¼ •ô#*¨3" (instances in which ^%¼ •ô-#e3s are used)
æþ – n"u$'y#*¨3" (conjugations in n"u1 $'y*& #390)
æ – $'y$T|#y *¨3" (when $T"y is presented as $'y)
æ – i$"%"‡ y#*¨3" (the meanings of i$"%s)
5. $| ot#$%&' (
(
In this section, $| T-#e3s, as well as some related topics such as Q¹$"3ys and *+$"3ys
are taught. Sections are made only by QU"j"*3-¨', the order of QU"j"3/ number.
(
ææ – $| ot1 $| e#*¨3" (in $| T-#e3 section, $| e-#e3s taught from 3.1.93 to 3.1.132)
(
y | ot' (from
æ – ^ku$ 3.1.133 to 3.2.end, the end of the 2nd ^"o of the 3rd chapter)
æ! – ƒ&"o35 (3.3.1, 2)
(
æ" – ƒ§%$| ot' (from 3.3.3 to 3.4.76, where i$"% starts)
6. $"%$*un*´#$%&' (
In i•*6 }"t$…'o6 /, $"%$ and *un*$ are not taught elaborately. Only a few sūtras in
both topics are taught in this one section. For a student of Vedanta, this much is not
enough. All the sūtras in the topic of both $"%$ (from 1.4.23 to 55) and *un*$ (2.3.all, the
entire 3rd ^"o of the 2nd chapter) can be studied in either * }"t$…'o6 / or $"*2$".
æ# – *un(‡"y5 (The meanings of nominal case endings)
61
7. '" -#$%&' (
In QU"j"3/, '" is taught in 2.1.beginning to 2.2.end. i•*6 }"t$…'o6 / follows the
QU"j"*3-¨'. In this '" -#$%&, related topics such as modification of ^ku^y o in the ƒ§%ô1
Q*q$"% in 6.3.section, the 3rd quarter of 6th chapter, and #e3s at the end of '" in the
'" "t"5 section at the end of 5.4, the 4th quarter of 5th chapter, are taught.
æ$ – '" "5 ( '" in general, and ^6 -( ^"-
6 '" , taught from 2.1.1 to 2.1.4)
æ% – QZ3/n"u5 (from 2.1.5 to 2.1.21)
6
þ – T°7s5 (from 2.1.22 to 2.2.22)
– RºÔ/*‹5 (from 2.2.23 to 2.2.28)
– Ò)5 (2.2.29)
æ – '" "t"5 (from 5.4.68 '" "t"5 to 5.4.end)
8. T*}T-#$%&' (
T*}T-#e3s are taught from 4.1.76 to 5.4.end, in almost the entire two chapters.
i•*6 }"t$…'o6 / follows QU"j"3/ order.
– T*}T"5 (T*}T in general, ç„*0y$ (general) #e3s are taught)
! – Qê"*q$"%5 (descendant of …)
" – %´"±‡ y$"5 (from 4.2.1 to 4.2.66)
# – S"T%6 *‡ y$"5 (from 4.2.67 to 4.2.91, the place of …, etc.)
$ – 2•*s$"5 (any other meanings than above)
% – *u$"%"‡ y$"5 (modification of …)
!þ – l0*q$"%5 (from 4.4.1 to 4.4.74)
! – 3o*q$"%5 (from 4.4.75 to 4.4.end)
! – ¶3T9›*q$"%5 (from 5.1.1 to 5.1.17)
!æ – lÈ*q$"%5 (from 5.1.18 to 5.1.114)
! – •Ti9%*q$"%5 (the status of …)
!! – nu4"±‡ y$"5 (from 5.2.1 to 5.2.93)
!" – '•‡*3"5 (one who has …)
!# – #"*Ïo2/3"5 (`, T , ( ‡", etc.)
!$ – #"*0u/3"5 (comparative, superlative)
62
Introduction
y "5 (T*}T#e3s which do not modify the meaning)

!% – €"*‡$
9. ¥/#e3-#$%&' (
Some #"*T^*o$s, in the sense of feminine, take ¥/#e3s. This section teaches all the
seven ¥/#e3s with conditions in which they are applied.
"þ – ¥/#e35 (from 4.1.3 to 4.1.81)
10. Appendices
k / – The index ( S
Q‡ ^*%*2U*us3 S k /) of the remaining (^*%*2U) subject (*us3)
" – *i¹^*%S35 (regulations on the gender of nouns)
" – Alphabetical index of `k s which obtain in i•*6 }"t$…'o6 /
"æ – Alphabetical index of u"*Ty$s which obtain in i•*6 }"t$…'o6 /
" – Alphabetical index of q"Ts6 which obtain in i•*6 }"t$…'o6 /
"! – Alphabetical index of q"Ts6 which were not introduced in i•*6 }"t$…'o6 /
"" – Index of 0&s which obtain in i•*6 }"t$…'o6 / in the order of appearance
"# – Alphabetical index of ^*%n"s"s which obtain in i•*6 }"t$…'o6 /
63
Layout of i•*6 }"t$…'o/

6
Each page of i•*6 }"t$…'o6 / consists of two parts: main part and *»å&/, footnote.
The main part consists of `k s and i•*6 }"t$…'o6 /u|*§, abbreviated as LSK in this book.
( bold)
`k ' (in
u|*§5 (short commentary)

written by u%o%"—"S"3y, the
autor of i•*6 }"t$…'o6 /
*»å&/ number
( brackets)
u"*Ty$' (in
*»å&/ (footnotes)
written by editor at
Gita Press
64
Introduction
Goals of studying i•*6 }"t$…'o/

6
The student should be able to explain:

• What the sūtra does
• Meaning of each word in the sūtra and anuvItti
• The –^* *} orally as well as in writing
• The topic of the sūtra in i•*6 }"t$…'o6 /
• The topic of the sūtra in QU"j"3/
Taking notes following i•*6 }"t$…'o/

6
The sūtras are to be studied one by one in the order of i•*6 }"t$…'o6 /. One sūtra is to
be written per notebook page, with the following items detailed for each sūtra:
• Type of the sūtra (given by n •'/Z"¤" or by the teacher)
• Sūtra number and the sūtra (with sandhi)
• Q4u|6 *§5 (words taken from previous sūtras)
• ô•1o5 (resolution of sandhi), and case and number of each word of the sūtra and
Q4u|6 *§5
• #"*T^*o$, resolution of '" , $| ot, and T*}T"t (if any), and meaning of *un*´ of each
word
• Meaning of the sūtra
• Meaning of the LSK u|*§5
• (
ƒo"‹%&' (examples) with step by step –^* *} with sūtras
The student is expected to prepare the notebook beforehand as much as possible.
In the i•*6 }"t$…'o6 / book, write sūtra numbers (in Arabic font) in front of the sūtra,
and Q4u|6 *§5 after the sūtra, so that the number and meaning of the sūtra can be easily
understood and memorized.
E.g., “6.1.77 p$9 3&*S «"« «##« ~ *‹T"3"'”(
65
Prayer of i•*6 }"t$…'o/

6
The j"4+9$ of Sarasvatī prepares our mind for the study.
j"4+9$5
6 " ÚÛ*uS"% "%^%'"'"±" —0o (Z"*^4-
2,
6
u/&"^Æ$q"*%&/'n3o" —"."+$"%"^‹"' «(
‹Æ1 ê"*»$'"*i$" *uoqT- ^/" 4 1 *vT"
(
u¡1 T" ^%'1Â%- n0uT- R*6 }#o" 2"%o"' -
6 "' 2/1
2, ( w±"' 2/1
( ÚÛ*uS"% "%^%'"' 2/1 ( —0o (Z"*^4/' 2/1
(
6
u/&"^Æ$q"*%&/' ( Qn3o"' 2/1
2/1
( —"."+$"%"^‹"' 2/1
( «
( *uoqT/' 2/1
‹Æ1 7/1 ê"*»$'"*i$"' 2/1 ( ^/" 4 1 7/1 *vT"' 2/1
(
( ^%'1Â%/' 2/1
u¡1 I/1 T"' 2/1 ( n0uT/' 2/1
( R*6 }#o"' 2/1
( 2"%o"' 2/1
( -
In the sentence, the first thing to find is the verb. Here, (Q‹) u¡1 « - I salute.
To whom do I salute? All those 2nd case ending words are pointing out the object of
6 "' –( white
saluting, %€T/. Each 2nd case ending word describes %€T/ for visualization: 2,
in colour; ÚÛ*uS"% "%^%'"' –( the ultimate essence of the inquiry of Brahma; w±"' –( one who
exists even at the beginning; —00"*^4/' –( one who pervades the whole universe;
6
u/&"^Æ$q"*%&/' –( one who holds vī ā and the Veda; Qn3o"' –( one who gives fearlessness;
—"."+$"%"^‹"' –( one who takes away the darkness of ignorance; ‹Æ1 ê"*»$'"*i$" *uoqT/' –(
one who holds the crystal mālā in hand; ^/" 4 1 *vT"' –( one who sits in the seat of lotus; T"
^%'1Â%/' –( to that ultimate goddess; n0uT/' –( one who has all the bhaga, what we want;
R*6 }#o"' –( one who gives knowledge; 2"%o"' –( the goddess Sarasvatī.
66
Introduction
At the top of the first page is the prayer to *u11Â%.

-./0&12"3 4'5-
To study Z"$%&, we need a lot of grace. We have to keep praying for wj"*ö$,
w*qn…*T$, w*qo•*u$ 2"*t.
Q‡ i•*6 }"t$…'o6 /
The word Q‡ can be taken in the sense of the beginning, and '¹i"‡ y. Now,
i•*6 }"t$…'o6 / begins.
The meaning of the title i•*6 }"t$…'o6 / is as follows:
$65 p*T ^|*‡u/, '‹/ « (earth)
$… '9oT1 p*T $6'o6 5 « (one who rejoices on the earth, lotus)
TŠ $6'o6 Š p3 $…'o6 / « (that which is meant for lotus, moonlight)
u"*o#*Tu"*o8" *4&*T5 Qt5 Qu "4 31s" T1 * }"t"5 «
* }"t"4" $…'o6 / pu $…'o6 / p*T * }"t$…'o6 / « (IB)
i2/ S" … * }"t$…'o6 / p*T i•*6 }"t$…'o6 /« (KT)
Now comes the prayer verse composed by ./u%o%"—"S"3y, the author of i•*6 }"t$…'o6 /.
6 $%9Å‹' «(
6 " 0Á3"
4•" %€T- o1u- 2}
(
^"*&4/3#u12"3 i•*6 }"t$…'o6 /' -
( o1u/' 2/1
4•" 0 %€T/' 2/1 ( 2} ( 0Á3"'
6 "' 2/1 6 ( $%9*' I/1 Q‹' 1/1
2/1
( «
( -
^"*&4/3#u12"3 4/1 i•*6 }"t$…'o6 /' 2/1
The meaning is given in the *»å&/ 1.

( u%o%"—"S"3y5 1/1 2}
[LSK *»å&/ ] Q‹' 1/1 ( o9s%*‹T"' 2/1
6 "' 2/1 ( 0Á3'
6 ( #2Æ0&3
2/1
6 ´ (
6 "' 2/1 (
%€T/' 2/1
( 4•" 0 4'-| e 0 ^"*&4/3#u12"3 4/1 R"i"4"' 6/3
o1u-' 2/1 u"Ïo1uT"' 2/1 ( ^"*&4/3Z"$%&2"¥1 7/1 #u12"‡'y 0(
( $%9*' I/1 p*T 0 Qµ35 1/1 -
i2/ 1/1 S 0 Q … 1/1 * }"t$…'o6 / 1/1 T"' 2/1
67
Q‡ !"-#$%&' (
In the very beginning, all the letters of the Sanskrit language have to be introduced.
[ !" `k "*&] 1. Qpƒ& «( 2. x{$ ( « 3. ‘,™ ( « 4. ¾çS «( 5. ‹3u%» ( «

6. i• & «( 7. È'™&4' «( 8. ðnÈ «( 9. •ñqs «( 10. —R0¸o2 «(
11. ÓÝ¶l‡S»Tu «( 12. $^3 «( 13. 2s % «( 14. ‹i ( -
[LSK] p*T 0 '"‹1Â%"*& 1/3 `k "*& 1/3 «

These are called '"‹1Â%"*& `k "*&.
(
That which came from '‹1Â% is called '"‹1Â%. '‹1Â%"T w0T"*4 p*T '"‹1Â%"*& « by 4.3.74
TT w0T5 «, '‹1Â% + ™* • + Q&. ( Why are they called by the name of '‹1Â%? *»å&/ 2 shows two
ślokas depicting the story of how Pā ini received these fourteen sūtras from Śiva.
[LSK *»å&/ ]
4|§"u "4 1 4»%"—%"—9 44"o ¸4" 4u^Íu"%' «(
(
y6 "'5 4$"*o* }"4 1T*Ò'2Ë *2u `k —"i' -
ƒ}T$
( 4u-^Íu"%' 2/1
4|§-Qu "4 1 7/1 4»%"—-%"—5 1/1 44"o III/1 ¸4"' 2/1 ( «
( ‘To (2/1 *u'2Ë I/1 *2u `k -—"i' 2/1
y6 "'5 1/1 4$"*o-* }"4 2/3
ƒ}T$ ( -
Only the first line is relevant to this topic now. At the end of the dance, the king
4»%"— sounded the damuru fourteen (9 + 5) times… In this śloka, '"‹1Â% `k "*& are called
*2u `k "*&. They are called #e"‹"% `k "*& as well, as they are meant for making #e"‹%.
314"“% '"Ä"3'*q0Å'‹1Â%"T «(
$| ¿ Z"$%& #9´ T¼ • ^"*&431 4'5 -
The meaning of this śloka is already introduced in this book.
Why are they called `k "*&? *»å&/ 3 shows the criteria for sūtra, as well as varieties of sūtra,
which we have already seen in the earlier part of this book. Since Q&, ( etc., !"s are made
from '"‹1Â% `k "*&, they are also considered to be !" `k "*&.
68
!"#$%&' (
The next word in u|*§ is an adjective to the preceding words, '"‹1Â%"*& `k "*&, to
explain what these sūtras are meant for.
[LSK] Q&"*o !"‡"y*4 1/3 «
(
These fourteen sūtras are meant for making Q& etc., !"s, which are known as
#e"‹"%s.
( w*o5 1/1 3" "' 6/3
Q& 1/1 ( T"5 1/3 Q&"o35 1/3 (116B)« Q&"o35 1/3 !"5 1/3 Q&"*o !"5 1/3 (KT)«
( 31s"' 6/3
Q&"*o !"5 1/3 Q‡5y 1/1 #39—4' 1/1 ( T"*4 1/3 Q&"*o !"‡"y*4 1/3 (116B) = '"‹1Â%"*& 1/3 `k "*& 1/3 «
( ( `k "&"' 6/3
[LSK] ‘s"' 6/3 ( ) Q•"5 1/3 pT5 1/3 «
The last letters of these sūtras are pTs.(

Qt1 nu"5 Q•"5 « That which exist at the end are said to be Q•"5.
[LSK] ‹$"%"*os 6 7/3 Q$"%5 1/1 ƒí"%&"‡5y 1/1 «
In the letters (consonants with Q added) starting from ‹ (5th sūtra), the letter Q is
meant for pronunciation, because the consonant by itself is difficult to pronounce. Thus, Q
is not intended.
( 5y #39—4 3Š 5 ƒí"%&"‡5y (116B)« For which ƒí"%&, pronunciation is the Q‡ y,
ƒí"%&' Q‡
purpose is called ƒí"%&"‡ y, adjective to Q$"%.
[LSK] iÁ'j1 7/1 T 6 0 pT-( !$5 1/1 «
( th sūtra) is pT. ( Q44"*
Whereas, the letter Q in the middle of i• & (6 6 ( 1.3.2
$ Q$"% is pT by
6
ƒô12›1 —44"* ( The purpose of this is for 1.1.51 ƒ%& %•( ^%5«
$ pT «.
Small questions:
How many sūtras are there in '"‹1Â% `k "*&?
How many letters are there in '"‹1Â% `k "*&?
69
Summary of this #$%&
1. Presenting all the letters by '"‹1Â% `k s

First, '"‹1Â% `k s are taught to present all the letters in Sanskrit.
There are only 42 letters pronounced in '"‹1Â% `k s. These 42 letters can represent all
the letters in the language by the device taught in the following sūtras.
2. Making #e"‹"%s
#e"‹"%s, abbreviations for certain groups of letters, are used as an important device
throughout QU"j"3/ `k s. #e"‹"%s are defined by the sūtra 1.1.71 w*o%•14 ‹1T" «.
pT, ( indicatory letter, is another important device in QU"j"3/ `k s. Since pT is
( used for
(
making #e"‹"%, the definition of pT has been taught beforehand. 1.3.3 ‹i•' «( is one of the
(
!" `k s for pT and (
1.3.9 TŠ i9^5« teaches how pT should be treated.
3. Varieties of sounds
One sound can have up to 18 varieties. These 18 varieties are categorized into three.
One is length, another is €%, intonation, and the other is nasal and non-nasal.
The length is three-fold. This is told by 1.2.27 ù$"i9› €o/•‚y 6T5«. The €% is also
6
three-fold, told by three sūtras: 1.2.29 ƒí •7o"T5«, 1.2.30 4/S •%4o"§5«, and 1.2.31 '"‹"%5 €*%T5«.
The nasal sound is defined by 1.1.8 'Ó 6
6 4"* $"uS49›44"* $5«.
4. Definition of u&y, similar letter
One letter can be a similar letter to another. This similarity of the letters is defined
(
6 "Š#3É u&y' «.
by 1.1.9 TÞ
5. Letters which represent u&ys ( u&y-œ"‹$)
One letter can represent its similar letters, u&ys. This is taught by the sūtra 1.1.69
Q&*6 o„u&yŠ S"#e35«. By this, the 42 letters told in '"‹1Â% `k s can represent all the letters in
Sanskrit.
70
!"#$%&' (
The next sūtra gives pT-( !" to the consonants at the end of each '"‹1Â% `k .
[ !" `k ']( 1.3.3 ‹i•' «( ~ ƒô12 1 pT (

In the teaching, the consonants at the end is pT. (
‹i 1/1 ( « ~ ƒô12 1 7/1 pT 1/1

( Q•' 1/1 (

• ‹i ( 1/1 – #e"‹"%5 ‹i.( All the consonants. This is !/.
• (
Q•' 1/1 – Qt1 nu5 p*T Q•' «( That which exists at the end is Q•'. ( This qualifies the
!/ ‹i.( Together, the consonants at the end.
• ƒô12 1 7/1 – #"*T^*o$ is ƒô12, the teaching taught by '*6 4`3'; ( in Q*q$%&1 â'/.
What is ƒô12 is explained in the u|*§.
• (
( ‹i 1/1
[LSK] ƒô12 1 7/1 Q•' 1/1 ( Š"T III/1
( pT 1/1 ( «
The consonant at the end in the teaching is termed pT. (
Now, what is ƒô12 is answered.
( «
[LSK] ƒô125 1/1 w±9í"%&' 1/1
ƒô12 is what is taught at the beginning (original teaching).
They are namely the following:
#e3"5 *2u `k "*& ã"o12" w0'"Æ‡" «
q"T^6 "l9 0&1 ^"l ƒô12"5 #$/*TyT"5 -
1) #e3"5 Suffixes taught in sūtras. E.g., 4, ( T|S, ( Q& (
2) *2u `k "*& '"‹1Â% `k "*&. E.g., Qpƒ& «( x{$ ( « ‘,™ ( «
3) wo12"5 Substitutes taught in sūtras. E.g., Qu™,( ‘2 (
4) w0'"5 6 ,( 4'6 (
Augments taught in sūtras. E.g., p» (, T$
5) q"T^6 "l5 (
Those which are enumerated in q"T^6 "l5. E.g., 2/™ ( €51 «, 6È ‹%&1 «
6) 0&^"l5 Those which are enumerated in 0&^"l5. E.g., w™ ( «
( ô' 1/1
[LSK] `k s1 6 7/3 Q”U' 1/1 ( ( Q4uTy
`k "t%"T 5/1 (
6 4/3' 1/1 uy` 0 -
Q4u|6 *§ is explained; unseen words in sūtras should be brought from another sūtra
wherever needed.
71
This is a side note:

One may object that this 1.3.3 ‹i•' «( sūtra depends on #e"‹"% ‹i ( which is defined
(
by 1.1.7 w*o%•14 ‹1T" «, in which pT letter is used. Thus there is a mutual dependency,
QÕ9Õ".3, between the two sūtras.
• 1.3.3 ‹i•' «( requires #e"‹"% in order to define pT (
• ( order to define #e"‹"%
1.1.7 w*o%•14 ‹1T" « requires pT in
To resolve this QÕ9Õ".3 problem, 1.3.3 ‹i•' «( can initially be looked at in a
different way.
‹i ( in 1.3.3 ‹i•' «( can be taken as the last sūtra of '"‹1Â% `k . In that case, the
( Š"T
meaning of the sūtra becomes: “‹i”( p*T `k 1 Q•' pT ( ( The last letter in “‹i”( sūtra is
«.
termed pT. ( At this stage, only one pT letter
( “i”( has been made.
Then, by 1.1.7 w*o%•14 ‹1T" «, the #e"‹"% ‹i ( is achieved with the w*o letter ‹ ( and
(
Q• pT letter i,( and it indicates all the consonants. At this stage, only one #e"‹"% “‹i”( has
been made.
Now, with the #e"‹"% ‹i,( we can come back to 1.3.3 ‹i•' «( again to give pT-( !" to
all the consonants at the end. Then we can also come back to 1.1.7 w*o%•14 ‹1T" « again to
make all kinds of #e"‹"%s.
72
!"#$%&' (
After defining pT, ( it should be told that the pT is

( to be elided. First, the term for
elision is to be defined by the next sūtra.
[ !" `k ']( 1.1.60 Qo2y4 i9^5 «

Elision is termed i9^.
( i9^5 1/1 «
Qo2y4' 1/1
• (
Qo2y4' 1/1 ( Þ»6 ( = o2y4; 4 o2y4' p*T
– ”2 + ( Qo2y4'. ( This is !/.
• i9^5 1/1 – This is !".
( i9^ !' 1/1

[LSK] # ´Š 6/1 Qo2y4' 1/1 ( Š"T III/1
( «
The disappearance of an existing thing is termed i9^.
# ´ means that which is applicable for the operation of disappearance.
Why the extra word “# ´Š” is added by the u|*§$"%?

(
Because even a non-existent thing is also not seen, Qo2y4. For example, a $| T-#e3 *ò^ (
( not seen after o*q, it cannot be said
never existed after a #"*T^*o$ o*q. Even though *ò^ is
6 ( w0' by
that “there is a disappearance, i9^, of *ò^”.( If this were allowed, it would bring T$
6 ( «, which is Q*4U, not desired.
6.1.71 •€Š *^*T $| *T T$
73
(
The next sūtra tells that the letters termed pT get i9^, elided.
[*u*q `k ']( 1.3.9 TŠ i9^5 «

There is elision for that (pT).(
TŠ 6/1 i9^5 1/1 «

• TŠ 6/1 – #"*T^*o$ is To (, a pronoun, which refers to something under discussion.
Here, the discussion is about pT, ( which has been defined from 1.3.2 to 1.3.8.
Thus this TŠ 6/1 is understood as pT5 6/1.
• i9^5 1/1 – This is wo125.
(
[LSK] TŠ 6/1 pT5 6/1 i9^5 1/1 Š"T III/1 «
There should be the i9^ substitute (elision) for that pT. (
By this sūtra, all the last consonants of '"‹1Â% `k , being pT, ( are to be elided. Then one
may question the purpose of their being. The next u|*§ answers this doubt.
(
[LSK] &-wo35 1/3 ( 5 1/3 -
Q&-w±‡"y
(
Those consonants starting from & are meant for #e"‹"%s, which start from Q&. (
Small questions:
(
Identify the pT letter of each '"‹1Â% `k .
74
!"#$%&' (
The next sūtra defines #e"‹"%.
[ !" `k ']( 1.1.71 w*o%•14 ‹1T" « ~ €Š

(
The letter with pT letter at the end is a !" for its own and following letters up to the
(
pT letter in '"‹1Â% `k .
w*o5 1/1 Q•14 3/1 ‹ 0 pT" 3/1« ~ €Š 6/1

• w*o5 1/1 – The beginning letter. This is !".
• Q•14 3/1 – #"*T^*o$ is Q•. Qt1 nu5 Q•5« That which exists at the end is Q•. This is
adjective to pT".
• ‹ 0 – with
• pT" 3/1 – #"*T^*o$ is pT; ( in ‹-T|T/3". This goes along with w*o5.
• €Š 6/1 – This is !/.
( S 0 !" 1/1 Š"T III/1

[LSK] Q•14 3/1 pT" 3/1 *‹T5 1/1 ( ‹ 0) w*o5 1/1 €Š 6/1 'j0"4"' 6/3 ( «
(
The beginning letter along with the pT letter, which exists at the end, is a !" for
itself and 'j0s ('j1 0•*t p*T 'j0"5), those which go in between the beginning letter and
(
the last pT letter.
( p*T 0 Q-p-ƒ-u&"y4"' 6/3
[LSK] 3‡" 0 Q& 1/1 ( !" 1/1 «
For example, of “Q&”,( Q is w*o, the beginning letter. & is
( Q• pT. ( (& is
( found in Q p
(
ƒ & sūtra ( ‹i•').( The w*o Q, which is with Q• pT, ( is a !" for itself, Q and
which is pT by
'j0, which are p and ƒ.
( ‹i 1/1
[LSK] ‘u' 0( QS 1/1 ( Qi 1/1
( p*T 0 wo35 1/3 -
In the same manner, QS, ( ‹i,( Qi,( etc. #e"‹"%s are made by this sūtra.
As seen in the *»å&/, 42 #e"‹"%s made from '"‹1Â% `k "*& are found in QU"j"3/ `k ^"l
u"*Ty$.
Exercise:
For each #e"‹"%, identify how many letters are in it, and list the letters.
75
The next five sūtras show 18 types of sounds in 3 categories.

The first category is the length of the vowel.
[ !" `k ']( 1.2.27 ù$"i9› €o/•y‚T5

6 «
There are three types of lengths: 1 count, 2 counts, and 3 counts; each is termed •€,
o/•,y ‚ 6T, respectively.
( •€-o/•-y ‚ 6T5 1/1 «

ù$"i5 1/1 QS 1/1
• ù$"i5 1/1 – This is an adjective to QS. ( This is a compound of ù and $"i. The details
are explained in the u|*§.
• (
QS 1/1 – #e"‹%5QS; ( indicating all the vowels. This is !/.
• •€-o/•-y ‚ 6T5 1/1 – •€š o/•š
y ‚ 6Tš •€o/•‚y 6T5 (ID); the singular is wsy-#3905 (license of
x*ss). This is !".
[LSK] ƒš 1/1 ùš 1/1 ùæš 1/1 u5 1/3 (ID)«

A short ƒ, a long ù, and further elongated ùæ are together called u5. This is a
pT%1T%Ò) '" (ƒ + ù + ùæ = ù). Because of the u&yo/• y *+, even after putting all the letters
together, only one o/• y ù is the result of the sandhi.
( ù+Q =
u5 is the 1/3 form of #"*T^*o$ ù (ù + — = ( u =
( u5).
[LSK] u"' 6/3 ( ¨'"T 0( ‹€-o/•-y ‚ 6T- !5 1/1 Š"T III/1

( $"i5 1/1 pu 0 $"i5 1/1 3Š 6/1 5 1/1 QS 1/1 ( «
That QS, ( whose time (length) is like the time of ù (ƒ, ù, and ùæ), is termed ‹€5, o/•5y ,
and ‚ 6T5 respectively.
For example, Q is an QS, ( whose length is like the length of ƒ, thus termed •€5. In
the same manner, w is an QS, ( whose length is like the length of ù, thus termed o/•5y . wæ is
an QS, ( whose length is like the length of ùæ, thus termed ‚ 6T5.
The '" “ù$"i” is an “IB”, pu-RºÔ/*‹- '" 5.
( ‹€-o/•-y ‚ 6T- !5 1/1)
( $"i5 1/1 pu 0 $"i5 1/1 3Š 6/1 5 1/1 ù$"i5 1/1 (QÕô"‡ y is QS 1/1
u"' 6/3
( $"i + • 6 + $"i + • 6
ù + w' + (u"*Ty$')( âÅ^'"4^k
6 u^y oŠ9§%ôi9^š«.
ù + $"i 6 q"T#6 "*T^*o$395« ~ $ (
2.4.71 ^9
76
!"#$%&' (
ù$"i (one whose time is like ù, which is ƒ, ù, ùæ)

The LSK u|*§ gives an introduction to the next set of three sūtras.
( ) #e1$' 0( ƒo"§"*on1o41 3/1 *`q" 0 -
[LSK] 5 1/1 (QS 1/1
Each one of those vowels, which were given •€-o/•-y ‚ 6T- !"s, are three-fold by the
6
division of ƒo"§, Q4o"§, and €*%T.
[ !" `k ']( 1.2.29 ƒí •7o"§5 « ~ QS (

The vowel pronounced from the upper portion of the v"4 is termed ƒo"§5.
(
ƒí •5 0 ƒo"§5 1/1 « ~ QS 1/1
• ƒí •5 0 – (Pronouncing) from upper portion, high; qualifying QS. (
• ƒo"§5 1/1 – This is !".
• (
QS 1/1 – This is !/.
There is no symbol assigned to ƒo"§5 for grammatical purpose in grammar books.
[ !" `k ']( 1.2.30 6

4/S •%4o"§5 « ~ QS (
6
The vowel pronounced from the lower portion of the v"4 is termed Q4o"§5.
4/S •5 0 Q4o"§5
6 1/1 (
« ~ QS 1/1
• 4/S •5 0 – (Pronouncing) from bottom part, low; qualifying QS. (
• 6
Q4o"§5 1/1 – This is !".
• (
6
The symbol assigned to Q4o"§5 in grammar books is a horizontal line under the
vowel. Ex. ‘q7 u|}… «
77
[ !" `k ']( 1.2.31 '"‹"%5 €*%T5 « ~ QS (

6
The vowel which consists of both the properties of ƒo"§ and Q4o"§ is termed €*%T5.
(
'"‹"%5 1/1 €*%T5 1/1 « ~ QS 1/1
• 6
'"‹"%5 1/1 – Combination of the qualities of ƒo"§ and Q4o"§. This is adjective to QS. (
• €*%T5 1/1 – This is !".
• (
The symbol assigned to €*%T5 in grammar books is a vertical line above the vowel.
Ex. œ‹8 ƒ^"o4 1 «
The LSK u|*§ gives an introduction to the next sūtra.

( ) 4u*uq5 1/1 Q*^ 0 #e1$' 0( Q44"*
[LSK] 5 1/1 (QS 1/1 6 6
$•-Q444"* ( *Òq" 0 -
$•8"' 3/2
Each one of those vowels is nine-fold. Now, each becomes further two-fold as nasal
and non-nasal by the next sūtra.
[ !" `k ']( 1.1.8 6

'Ó4"* 6 $5 «
$"uS49›44"*
6
The sound pronounced by the nose along with mouth is termed Q44"* $5.
6 -4"* $"-uS45 1/1 Q44"*

'Ó 6 $5 1/1 «
• 6 -4"* $"-uS45 1/1 – This is !/. The sound pronounced by the nose along with mouth.
'Ó
This '" is resolved below.4
• 6
Q44"* $5 1/1 – This is !".
6 - *‹T-4"* $3" 3/1 ƒí"3y'"&5 1/1 u&y5 1/1 Q44"*

[LSK] 'Ó 6 (
$- !5 1/1 Š"T III/1 «
The sound which is pronounced by the nose along with the mouth is termed
6
Q44"* $.
4
6 41 3/1 *‹T" 1/1 'Ó
'Ó 6 *‹T" 1/1 (3T) « 'Ó
6 *‹T" 1/1 4"* $" 1/1 'Ó
6 *‹T4"* $" 1/1 (KT) « 'Ó
6 *‹T" 1/1 4"* $" 1/1
(
6 4"* $" 1/1 ('j'ôi9^/ type of KT, by u"*Ty$' 2"$^"*‡
'Ó ( « ƒÎT1 p*T
uy "o/4" * }31 ƒ§%ôi9^Š9^ ¤"4' «)
6
uS45 ($'y*& Z°*§5) 6 4"* $3" ƒí"3y'"&5 uS45 (3T)«
« 'Ó
78
!"#$%&' (
Now, LSK u|*§$"% summarizes the result of the previous five sūtras.
( pô' 0( – “Q p ƒ x” ‘s"' 6/3
[LSK] TT 1/1 ( u&"y4"' 6/3
( #e1$' 0( QU"o2 1/3 n1o"5 1/3 «
“TT”( indicates the result of the previous five sūtras. “pô'”( means “in this manner”.
There are 18 types for each letter Q, p, ƒ, x.
6
Q444"* $5 6
Q44"* $5
•€5 o/•5y ‚ 6T5 •€5 o/•5y ‚ 6T5
ƒo"§5 Q w wæ Q• w• w•æ
€*%T5 Q8 w8 w8æ Q• 8 w• 8 w• 8æ
6
Q4o"§5 Q7 w7 w7æ Q•7 w•7 w•7æ
Same for p, ƒ, 9.
( «
[LSK] {u&yŠ 6/1 (n1o"5 1/3) Ò"o2 1/3 TŠ 6/1 o/•-y Qn"u"T 5/1
Because of the absence of o/•,y there are 12 types of {.
6
Q444"* $5 6
Q44"* $5
•€5 o/•5y ‚ 6T5 •€5 o/•5y ‚ 6T5
ƒo"§5 { - {æ {• - {• æ
€*%T5 {8 - {8 æ {• 8 - {• æ8
6
Q4o"§5 {7 - {7 æ {•7 - {•7 æ
( Q*^ 0 Ò"o2 1/3 T1s"' 6/3

[LSK] ‘S"' 6/3 ( •€-Qn"u"T 5/1
( -
( ,, ¾, ç), because of the absence of •€.
There are also 12 types of ‘S (‘,
6
Q444"* $5 6
Q44"* $5
•€5 o/•5y ‚ 6T5 •€5 o/•5y ‚ 6T5
ƒo"§5 - ‘ ‘æ - ‘• ‘•æ
€*%T5 - ‘8 ‘8æ - ‘• 8 ‘• 8æ
6
Q4o"§5 - ‘7 ‘7æ - ‘•7 ‘•æ7
Same for ,, ¾, ç
79
Letters in Sanskrit are grouped by the similarity in pronunciation.

The next sūtra defines u&y, similar letter.
[ !" `k ']( 1.1.9 6

TÞ"Š#3É u&y' «(
Letters which have the same wŠ (= v"4) and #3É (= w8t%#3É) are grouped as
u&y.
(
6 -wŠ-#3É' 1/1
TÞ ( «
u&y' 1/1
• TÞ (
6 -wŠ-#3É' 1/1 – This is !/.
(
6 )1 nu' wŠ'
wŠ1 ('Ó ( « That which is being in the mouth is called wŠ.
5
wŠ here means v"4.

#$| U5 3É5 #3É5 (PT)« This indicates w8t%#3É5
6 … wŠ#3É… 3Š TT T( Þ
TÞ (
6 "Š#3É' (116B) «
• (
u&y' 1/1 – This is !". This can be translated as “similar letter”
[LSK] T" -w*o-v"4' 1/1 ( Ò3' 1/1

( w8t%#3É5 1/1 S 0 p*T 0 ‘TT 1/1 ( 3Š 6/1 (u&yŠ 6/1) 314 3/1 (Q^%1&
3/1
u&Ë4 3/1) TÞ ( TT 1/1
6 ' 1/1 ( (u&y5 1/1) *'‡5 0 u&y !' 1/1
( Š"T III/1
( «
Letters which have the same v"4 and w8t%#3É are termed u&y with reference to
each other.
y and { is o•.
According to this definition, x and { are not u&ys because x is 'kqÕ
However, to comply with how the language is, x and { have to be treated as u&y to each
other. Since Pā ini did not specifically mention the status of u&y belonging to these two
letters, u"*Ty$$"% makes a statement.
5
( 14 p*T wŠ' ($%&1
QŠ*t ƒí"%3*t u&"y4 Q4 ( 6
Z°*§5)
( Á3T (
Q + 3.1.124 x‹i9Á3yT «( 3.3.113 $| eÞ»9
6 Rºi' «(
wŠ
wŠ + 3T ( 4.3.55 2%/%"u3u"í« ~ T` nu5 3T (
( 3
wŠ + 6.4.148 3Š1*T S « ~ nŠ T*}T1 i9^5
wŠ 8.4.64 ‹i9 3'" 3*' i9^5 «
80
!"#$%&' (
(u"*Ty$')( x{u&y39*'y‡5 "uÁ3Ñ u"Î' «(

It has to be said that xu&y and {u&y are u&y to each other.
( u"Î' 1/1
x-{-u&y395 6/2 *'‡5 0 "uÁ3y' 1/1 ( «
4 words in the u"*Ty$, other words are understood by the context of the topic under
discussion.
• x-{-u&y395 6/2 – 18 types of x and 12 types of {; in ¬+sé/ to "uÁ3y' (
• *'‡5 0 – mutually, with reference to each other
• (
"uÁ3y' 1/1 (
– u&yŠ n"u5 "uÁ3y' (³È )( « The status of u&y, which belongs to the 6th case-
ending word, x-{-u&y395. In the translation, the 6th case and ³È can
( be cancelled and
put in apposition, resulting in “xu&y and {u&y are u&y”
• (
u"Î' 1/1 ( Á3T ($'y
– uS + ( *& with an added sense of necessity) « That which needs to be
said.
Now LSK u|*§$"% has to show what are wŠ (= v"4) and #3É.
First, v"4s of letters are told.
( (v"4' 1/1
[LSK] Q$6‹*u —y4/3"4"' 6/3 ( ) $Ál5 1/1 «
The v"4 of Q (Qu&y), $6 ($u0y letters which are $ ( Ó, ( 0, ( •, ( ™)( , and *u —y4/3 (= *u 0y) is
$Ál.
6
[LSK] pS32"4"' ( T"
6/3 1/1
«
The v"4 of p (pu&y), S 6 (Su0y letters which are S, ( ¶ (, —, ( ð, ( È),( and 2 is
( T" . (T" is an ƒ-
ending neuter word.)

( 'kq"y 1/1 «
[LSK] x»6%s"&' 6/3
The v"4 of x (xu&y), »6 (»u0y letters which are » (, l (, ¸ (, ñ,( &),( and s is
( 'kq"y. (#"*T^*o$ of
“'kq"y” is 'kq4y , ( masculine word)

( ot"5 1/3 «
6 "4"' 6/3
[LSK] {Ti
The v"4 of { ({u&y), T 6 (Tu0y letters which are T, ( ‡, ( o (, q, ( 4, ( and )( is ot"5. (Plural is
because of the plurality of teeth.)
( ,é… 1/2 «
[LSK] ƒ^k^:"4/3"4"' 6/3
81
The v"4 of ƒ (ƒu&y), ^ 6 (û0y letters which are ^, ( Ý,( R, ( n, ( '),( and ƒ^:"4/3 (the *u 0y after
(
^ and Ý)( is ,é…. (Dual is because lips are two.)
( 4"* $" 1/1 S «
[LSK] È'™&4"4"' 6/3
The v"4 of È, ( ', ( ™,( &, ( 4 is
( 4"* $". The S$"% (“S”) is because all these nasal sounds are
already introduced in the respective v"4. Thus, these nasal sounds have two v"4s each.
[LSK] ‘o•T95 6/2 $ÁlT" 1/1
«
( ¾T S
The v"4 of ‘ and ¾ is $ÁlT" , the combination of throat and palate. ‘T S ( ‘o•T…
(
(ID), T395 ‘o•T95. The T after ‘ and ¾ is for clear presentation of the letters. Without the T, ( all
letters are subject to sandhi rules, after applying which it is difficult to read.
( «
[LSK] ,o…T95 6/2 $Ál9é' 1/1
The v"4 of , and ç is $Ál9é', ( the combination of throat and lips.
( «
[LSK] u$"%Š 6/1 ot9é' 1/1
( ot9é', ( the combination of teeth and lips.
The v"4 of u is
( «
[LSK] *—;"'ki/3Š 6/1 *—;"'ki' 1/1
The v"4 of *—;"'ki/3 (the *u 0y after $ ( and Ó)( is *—;"'ki', ( the root of tongue.
[LSK] 4"* $" 1/1 Q4€"%Š
6 6/1
-
6
The v"4 of Q4€"% is 4"* $".
Next, two #3És are told.

[LSK] 3É5 1/1 *Òq" 0 – w8t%5 1/1 R"ã5 1/1 S 0 «
3É is two-fold; w8t% (internal) and R"ã (external).
Note that “internal” and “external” are just names. w8t%#3É is made before the
sound is heard, whereas R"ã#3É is made while making the sound.
Now, five types of w8t%#3É are told.

( «
[LSK] w±5 1/1 ^Íq" 0 – Ð|U- s<Ð|U- s*=uu|T-*uu|T- u|T-n1o"T 5/1
The first one, namely w8t%#3É is five-fold: 1. Contact, 2. Slight contact, 3. Slightly
open, 4. Open, 5. Contracted.
( #3É' 1/1
[LSK] T` 0 Ð|U' 1/1 ( Ð2"y4"' 6/3
( «
There, among w8t%#3É, the Ð|U #3É is of Ð2ys.
82
!"#$%&' (
Ð2ys are specified later in this section as:

[LSK] $-wo35 1/3 '-Qu "4"5 1/3 Ð2"y5 1/3 «
Letters starting with $ ( and ending (Qu "4) with ' ,( 25 consonants, are Ð2ys.
[LSK] s<Ð|U' 1/1 ( ) Qt5v"4"' 6/3

( (#3É' 1/1 ( «
The s<Ð|U#3É is of Qt5vs, semi vowels.

Qt5vs are specified later in this section as:
[LSK] 3&5 1/3 Qt5v"5 1/3 «
3&s( (3 ,( u ,( % (, i)( are Qt5vs.
[LSK] s*=uu|T' 1/1 ( ) ù &"' 6/3

( (#3É' 1/1 ( «
The s*=uu|T#3É is of ù 4s,( sibilants. ù 4 is

( a 4-ending
( masculine word.
ù 4s( are specified later in this section as:
[LSK] 2i5 1/3 ù "&5 1/3 «
2is( (2 ,( s ,( ,( ‹ () are ù 4s.(
[LSK] *uu|T' 1/1 ( ) €%"&"' 6/3

( (#3É' 1/1 ( «
The *uu|T#3É is of €%s, vowels.

€%s are specified later in this section as:
[LSK] QS5 1/3 €%"5 1/3 «
QSs( (all the vowels) are €%s.
( ) u|T' 1/1
[LSK] •€Š 6/1 Qu&yŠ 6/1 #3901 7/1 (#3É' 1/1 ( «
As for •€ Q, #3É is u|T in usage.

( T 6 0 *uu|T' 1/1
[LSK] #*¨3"o2"3"' 7/1 ( ‘u 0 «
However, in grammatical process, #3É of •€ Q is *uu|T.

To achieve u&yo/• y *+, •€ Q is treated as *uu|T sound while deriving a word. After
deriving, the •€ Q is converted to u|T sound by the last sūtra 8.4.68 Q Q «.
83
Now, explanation of R"ã#3É starts.

[LSK] R"ã#3É5 1/1 T 6 0 ‘$"o2q" 0 – *uu"%5 1/1 u"%5 1/1 Â" 5 1/1 4"o5 1/1 •9s5 1/1 Q•9s5 1/1 Q®#"&5 1/1
'‹"#"&5 1/1 ƒo"§5 1/1 Q4o"§5
6 1/1
€*%T5 1/1 S 0 p*T 0 «
R"ã#3É is eleven-fold.
[LSK] Ó%5 1/3 *uu"%5 1/1 Â" "5 1/3 Q•9s"5 1/3 S 0 «
Ó%s( (all the hard consonants) are called *uu"%, Â" , and Q•9s.
[LSK] ‹25 1/3 u"%"5 1/3 4"o"5 1/3 •9s"5 1/3 S 0 «
‹2s( (all the soft consonants) are called u"%, 4"o, and •9s.
( #‡'-T|T/3^Í'"5 1/3 3&5 1/3 S 0 Q®#"&"5 1/3 «
[LSK] u0"y4"' 6/3
1st, 3rd, and 5th of the classes and 3&s( are called Q®#"&.
( *ÒT/3ST‡6 > 1/2 2i5 1/3 S 0 '‹"#"&"5 1/3 «
[LSK] u0"y4"' 6/3
2nd and 4th of the classes and 2is( are called '‹"#"&.
These LSK u|*§s are already explained in the w8t%#3É section in this book.
[LSK] $-wo35 1/3 '-Qu "4"5 1/3 Ð2"y5 1/3 « 3&5 1/3 Qt5v"5 1/3 « 2i5 1/3 ù "&5 1/3 « QS5 1/3 €%"5
1/3
«
From here onward, the letters which are not seen in '"‹1Â% `k s are also explained.
( #"$ ( 0 Qqy-*u 0y- ”25 1/1 *—;"'ki/35 1/1«
[LSK] “: $ : Ó” p*T 0 $Ó"8"' 5/2
( called *—;"'ki/3.
The one which looks like half *u 0y before $ ( and Ó is
( #"$ ( 0 Qqy-*u 0y- ”25 1/1 ƒ^:"4/35 1/1«
[LSK] “: ^ : Ý” p*T 0 ^Ý"8"' 5/2
(
The one which looks like half *u 0y before ^ and Ý ( is called ƒ^è"4/3.
[LSK] “Q Q5” p*T 0 QS5 5/1 ^%… 1/2 Q4€"%-*u
6 0> 1/2 -
6
The dot above a vowel and two dots after a vowel are respectively called Q4€"% and
*u 0y.
84
!"#$%&' (
So far, all the letters used in Sanskrit language have been presented systematically,
and the concept of u&y has also been presented.
The next sūtra teaches that some letters in '"‹1Â% are !"s for their u&ys.
[ !" `k ']( 1.1.69 6 ( u&yŠ S"#e35 « ~ €Š

Q&*oT
(
Q& which is not ordained (as wo12, w0', or #e3), and ƒ*oTs( are !"s for their u&ys.
( ƒ*oT 1/1
Q& 1/1 ( u&yŠ 6/1 S 0 Q#e35 1/1 « ~ €Š 6/1
• (
Q& 1/1 – #e"‹"% Q&. ( This is !".
(
#e"‹"% Q& here with the 2nd &, ( ^%-&$"%, among pT letters
( in '"‹1Â% `k s.
• (
ƒ*oT 1/1 – This is also !".
( 3Š
ƒT pT ( (
5 ƒ*oT (116B) (
« The one whose pT letter is ƒ is called ƒ*oT. ( Viz., $6, S,6 »6, T,6 ^.6
• u&yŠ 6/1 – This is !/.
• (
S 0 – connecting Q& and ƒ*oT. (
• Q#e35 1/1 – This is an adjective to Q&. ( See LSK u|*§ for details.
• €Š 1/1 – This is also !/.
In this sūtra, “#e3” is used in the sense of “what is ordained”, not as we know in
the sense of suffix. The word #e3 is grammatically explained.
[LSK] #T/3T1 III/1 *uq/3T1 III/1 p*T 0 #e35 1/1
«
( QS. ( This $| T-#e3
The word “#e3” is comprised of #*T + p& + ( (
QS can be in the sense
of q"T 6 itself (n"u1) and any $"%$ other than $Ty*%. We need to know in which sense this suffix
( i» (/$'y*&/III/1) limits the sense of $| T-#e3
is used. The first word “#T/3T1” (#*T + p& + ( QS (
to $'y*&, by using the same ƒ^ 0y and q"T 6 with different #e3 which is *T™ ( in $'y*&. From
this, we understood that “#e3” is an object of #*T + p&. (
Now, the meaning of the combination of #*T + p&, ( which has many possible
meanings, has to be ascertained. The second word “*uq/3T1” (*u + q" + i» (/$'y*&/III/1) limits
the meaning of #*T + p&. ( The common meaning between #*T + p& and
( *u + q" is “to ordain,
to command”.
85
Now we understood that “#e3” means “what is ordained, what is commanded”,

namely, wo12, w0', and #e3.
Now the meaning of the whole sūtra is explained.

[LSK] Q*uq/3'"45 1/1
(= Q#e35) Q& ( 1/1 ƒ*oT ( 1/1 S 0
u&yŠ 6/1
!" 1/1
Š"T ( III/1 «
Q*uq/3'"4 Q& ( and ƒ*oT are !" for u&y.
( (/$'y*&) is a synonym of “Q#e3” (4È +
( *u + q" + 2"4S-i»
“Q*uq/3'"4” (4È + (
( #*T + p& +
( *&), which means “what is not ordained”, namely, any word indicating other than
QS-$'y
wo12, w0', or #e3. Q*uq/3'"4, or Q#e3, is an adjective to Q& .(
( this sūtra is explained.

The #e"‹"% Q& in
( ^%1& 3/1 &$"%1& 3/1 «
[LSK] Q` 0 ‘u 0 Q& 1/1
Only in this sūtra, #e"‹"% Q& ( is made with the 2nd & as
( pT in
( '"‹1Â% `k s, representing
all the vowels (Q, p, ƒ, x, {, ‘, ,, ¾, ç) and semi vowels (‹ (, 3, ( u, ( %,( i).(

( any other sūtra is with the 1st & as
That means Q& in ( pT in
( '"‹1Â% `k s, representing
only Q, p, ƒ.
( always with ^ku&
See the *»å&/ – Q& is y $"% by default, except for this sūtra 1.1.69
( always with ^%-&$"% without exception.
Q&*6 o„u&yŠ S"#e35«. p& is
( this sūtra is explained.

ƒ*oT in
[LSK] $• 6 S • 6 »• 6 T • 6 ^ • 6 - ‘T1 1/3 ƒ•*oT5 1/3 «
These $,( S, ( » (, T, ( ^ have
( (
ƒ• as pT letter. Thus these five letters are called ƒ•*oTs.( ƒ•*oTs(
are only these five.
(
6 "Š#3É u&y' «,
As per 1.1.9 TÞ u&ys of $ ( are Ó, ( 0, ( •, ( ™,( since they have the same
v"4, which is $Ál, and the same w8t%#3É, which is Ð|U. By this sūtra 1.1.69 Q&*6 o„u&yŠ
( • 6 is a !" for its u&y Ó, ( 0, ( •, ( ™,( and itself, by taking Q4u|6 *§ “€Š”. In the
S"#e35«, the ƒ*oT $
same manner, other ƒ*oTs( are !" for their u&ys and themselves.
(
Now Q*uq/3'"4 Q& as !" for its u&y is explained.
( ‘u' 0( – Q 1/1 p*T 0 QU"o2"4"' 6/3
[LSK] TT 1/1 ( !" 1/1 «
86
!"#$%&' (
Q, the first letter of Q&, ( is !" for 18 letters (17 u&ys and itself).
Because all 18 types of Q (3 lengths x 3 accents x 2 nasal/non-nasal) have the same
v"4, which is $Ál, and w8t%#3É, which is *uu|T, they are all u&ys to each other.
By this sūtra, the Q in '"‹1Â% `k is a !" for itself and all the 17 u&ys.
[LSK] T‡" 0 p$"%-ƒ$"%… 1/2 «

In the same manner, p and ƒ are !"s for 18 types of p and ƒ, respectively.
[LSK] x$"%5 1/1 *`2T5 6/1 «

x is !" for 30 letters.
(
Because of the u"*Ty$' “x{u&y ( the u&y of x is both 18 types of x
39*'y‡5 "uÁ3Ñ u"Î' «“,
and 12 types of {. Thus x in '"‹1Â% `k is a !" for 30 letters.
[LSK] ‘u' 0( {$"%5 1/1 Q*^ 0 «
For the same reason, { is !" for 30 letters.
( «
[LSK] ‘S5 1/3 Ò"o2"4"' 6/3
(
‘S are !"s for 12 letters each.
6
[LSK] Q44"* 6
$-Q444"* $-n1o41 3/1 3ui"5 1/3 *Òq" 0 « T14 3/1 Q444"*
6 $"5 1/3 (3ui"5 1/3) T1 1/3 Ò395 6/2
Ò395 6/2 !"5 1/3 -
3, ( u, ( i ( are two-fold by the division of nasal (3,• ( u,• ( i• )( and non-nasal (3, ( u, ( i).( Because
of this, these non-nasal 3, ( u, ( i ( in '"‹1Â% `k s are !"s for two letters each.
This sūtra is known as u&yœ"‹$- `k because it makes the letters in '"‹1Â% `k s grab,
mention, represent their u&ys.
Up to this point is the topic of u&y-*4oË25, presenting the letters.
87
From here, three !"s which are needed to proceed for the next section are
introduced.
The next sūtra defines *‹T", which is the topic of the next *+ section.
[ !" `k ']( 1.4.109 ^%5 *ü$s y5 *‹T" «

The most (closest) proximity of letters is termed *‹T".
^%5 1/1 *ü$sy5 1/1 *‹T" 1/1 «

• ^%5 1/1 – most, highest, maximum; this is adjective to *ü$sy5.
• *ü$sy5 1/1 – Nearness, proximity, vicinity; this is !/.
'+ ( come near + •È (n"u1
( *4 + $| s to ( )
• *‹T" 1/1 – This is !".
( *4 + q" to place near, to put together + ´ ($'y*&)
'+
( Q*T2*3T5 1/1 (= ^%5 1/1) *ü*q5 1/1 (= *ü$sy5 1/1) *‹T" !5 1/1 Š"T III/1
[LSK] u&"y4"' 6/3 ( «
The maximum closeness among letters is termed *‹T".
( *ü*q
( *4 + q" to place near, to put together + *´4 =
'+
In R"i'49%'" it is said that the gap between two letters should be less than a half '"`"
to be called ^%5 *ü$sy5.
88
!"#$%&' (
The next sūtra defines 390, conjunct consonants.
[ !" `k ']( 1.1.7 ‹i9›4t%"5 3905 «

Consonants which are not intervened by vowels are termed 390.
‹i5 1/3 Q4t%"5 1/3 3905 1/1 «

• ‹i5 1/3 – #e"‹"% ‹i,( all the consonants; this is !/.
• Q4t%"5 1/3 – This is an adjective to ‹i5.
(
Q*u±'"4' Qt% (= Zuq"4) 31s" T1 Q4t%"5 u&"y5 (NB)«
• 3905 1/1 – This is !".
[LSK] Q* n5 3/3 QZu*‹T"5 1/3 ‹i5 1/3 390 !"5 1/3 Š56 III/3 -
Consonants which are not intervened by vowels are termed 390.
*u + Qu + q" to separate, interrupt, obstruct + ´ ($'y*&) = Zu*‹T
4 Zu*‹T"5 QZu*‹T"5 (NT) « Not obstructed
In the sūtra, “Q* n5” (#"*T^*o$ is QS)( is not heard. Still, it is implied by “Qt%'”(
(
which means Zuq"4' (obstruction, partition). That which obstructs, intervenes the
consonants should be *u—"T/3 (that which belogs to another class) to consonants. Those
which are *u—"T/3 to consonants should be vowels. That is why those which intervene are
told as “QS”( in $Ty*% T|T/3".
89
The next sūtra defines ô, word used in Sanskrit language.
[ !" `k ']( 1.4.14 6

*â™t ô' «(
( *T™ ( is termed ô.
That which ends with ^6 or
(
^6 -*T™ ( ô' 1/1
-( Qt' 1/1 ( «

• (
^6 -*T™ (
-( Qt' 1/1 ( *T™ ( S ^6 -*T™…
– ^6 S ( (
(ID) « ^6 -*T™… Qt…/Qt1 3Š TT ( ^6 -*T™
( -( Qt' (
(116B); this is !/.
• (
ô' 1/1 – This is !".
6
[LSK] Rt' 1/1
( S 0 ô !' 1/1
( *T™t' 1/1 ( Š"T III/1
( -
(
That which ends with ^6 -#e3, and that which ends with *T™-( #e3 are termed ô.
( «
[LSK] p*T 0 !"#$%&' 1/1
Thus ends the topic of !".
90
Qˆ *+5
Q‡ QS-( *+5
(
*+ is a synonym to *‹T", the closest proximity of letters (u&"y4"' Q*T235 *ü*q5), as
we saw in the *‹T"- !"- `k .
*+ also means a phonetic change when two sounds meet without a gap.
When to apply *+ is told in the next śloka:
*‹T" ‘$ô1 *4e" *4e" q"Tk^ 0y395 «
'" 1 *4e" u"À1 T 6 " *uu“"'^1“T1 -
1) *‹T" 1/1 ‘$ô1 7/1 *4e" 1/1 «
*+ is compulsory within a word. E.g., n9 + Q + *T = nu*T
2) ( *‹T" 1/1) q"T-6 ƒ^ 0y395 7/2 *4e" 1/1 «
*+ is compulsory between ƒ^ 0y and q"T.6 E.g., *u + ‘*T = Z1*T
Technically speaking, ƒ^ 0y is considered to be an independent ô.
3) ( *‹T" 1/1) '" 1 7/1 *4e" 1/1 «
*+ is compulsory within a compound.
6
Compound, '" , is a word made of many words. A group of Rt-ôs gains '" -
!". In other words, words are compounded.
E.g., 6 ƒô125 0–ô1
0%95 6 25 « This is i…*$$*uœ‹u"À.
( ƒô12 + • 6
= 076 + ™ + This is Qi…*$$*uœ‹u"À.
'" - !" 2.2.8 sé/ « ~ ^6 ( ^"
6 ‹ T°7s5
6 '" 5
This group of words gains '" - !".
#"*T^*o$- !" 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
By being '" , #"*T^*o$- !" is given.
= 076 + ƒô12 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
(
By being #"*T^*o$, all ^6 -#e3s are elided.
6
= 0–ô1
2 6.1.101 Q$5 u&Ë o/•5y « ~ Q*S ‘$5 ^ku^y %395 *‹T"3"' (
Between the two ôs in '" , *+ is applied.
( Q^1“T1 III/1 «
4) u"À1 7/1 T 6 0 " 1/1 ( *‹T" 1/1) *uu“"' 2/1
However, in a sentence, it depends on the speaker’s intention.
91
In i•*6 }"t$…'o6 /, *+ is divided into 3 types: QS-( *+, ‹i-( *+, and *u 0y *+.
QS-( *+ is a phonetic modification on vowels. Regulations on QS-( *+ are generally
( making the topic of *‹T"
found in *u*q `k s told in the section starting with 6.1.72 *‹T"3"' «,
in *us3 â'/.
Qˆ *+5 1) 3&-( *+5 “ ‰

6 ^"Š5”
6
In i•*6 }"t$…'o6 /, sūtras are presented through the process of –^* *}, the achieving
of a desired form of a word. A form (–^) is given as a theme first, then to achieve (* *})
that form, all the required sūtras are presented step by step.
For 3&-( *+ topic, the first theme given is “ q/
6 + ƒ^"Š”.
6
This is a part of the process of making a T|T/3"T°7s 6 is a masculine word,
'" . q/
6 (116B) «. ƒ^"Š is an object of worship or
meaning ^*Á¸T (scholar): 29n4" q/5 3Š 5 q/5
( Á3T ($|
meditation: ƒ^"ŠT1 p*T ƒ^"Š5 « ƒ^ + w + ( e) «
6
q/*n5 6 ^"Š5
ƒ^"Š5 p*T ‰ 6 « This is i…*$$*uœ‹u"À.
6 ^"Š.
The object to be worshipped by contemplative people is called ‰ 6
6 + »" + ƒ^"Š + • 6
q/ This is Qi…*$$*uœ‹u"À.
'" - !" y %&1 $| T" Rºi' «( ~ T|T/3" T°7s5
2.1.30 $T|$ 6 '" 5
#"*T^*o$- !" 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
6
q/ + ƒ^"Š 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
Now, between these two ôs in '" , *+ is applied.
92
Qˆ *+5
The next *u*q `k gives the “3&-( *+ rule”, as in the diagram below.
6 7
p$ ( QS (
1
3& (
[*u*q `k ']( 6.1.77 p$9 3&*S« ~ *‹T"3"' (

( the substitute in the place of p$ ( when it is followed by QS, ( in the topic of *‹T".
3& is
( Q*S 7/1 « ~ *‹T"3"' 7/1

p$5 6/1 3& 1/1 (

• p$5 6/1 – #e"‹"%5 p$;( in v"431 90" sé/ (1.1.49 sé/ v"431 90" «, as explained below)
• (
3& 1/1 – #e"‹"%5 3&; ( this is wo125.
• ( ^% â'/ (1.1.66 T*¼*ü*T *4oËU 1 ^kuŠ
Q*S 7/1 – #e"‹"%5 QS in y «, as explained below)
• (
*‹T"3"' 7/1 (
– The Q*q$"% `k 6.1.72 *‹T"3"' «; *‹T" is a !" defined as 1.4.109 ^%5
*ü$sy5 *‹T" «; in *us3 â'/.
( Š"T III/1
[LSK] p$5 6/1 v"4 1 7/1 3& 1/1 ( ( *us31 7/1-
Q*S 7/1 *‹T"3"' 7/1
In the place (v"4 1 7/1) of p$ ( (p$5 6/1), there should be (Š"T III/1 ( , when QS (Q*S
( ) 3& 1/1 (
7/1
) follows, in the topic of (*us31 7/1) proximity of sounds ( *‹T"3" 7/1).
The *T™t word “Š"T”( is supplied to complete the sentence. The word in 1st case is
( understood as wo12, a
always understood as $"3y, the change to be made. Thus 3& is
substitute. The word “*us31” is supplied to show the 7th case of “ *‹T"3"'”( is *us3 â'/.
93
Now questions arise: Where did the word “v"4 1”, which is connected to the 6th case
ending word, come from? How do we know that the “Q*S” in 7th case means “when QS (
follows”? These interpretations of the *un*´ are not the meanings of *un*´ as per natural
language. We need special meanings of *un*´ for reading sūtras. The following ^*%n"s" `k s
give special interpretations for *un*´s used only in sūtras.
The next ^*%n"s" `k is not taught in i•*6 }"t$…'o6 /.
[^*%n"s" `k ']( 1.1.49 sé/ v"41390" «

The 6th case, whose ¬+ is not understood, should bring the word “v"4 1”,
and connect with the 6th case ending word.
This kind of 6th case is called ^"*%n"*s$-sé/, or can be called v"431 90" sé/.
sé/ 1/1 v"4 1-390" 1/1 «

• sé/ 1/1 – This is ƒ Ÿ1 3, subject.
• v"4 1-390" 1/1 – This is *uq13, predicate.
“v"4”1 p*T ô14 3905 3Š"5 " v"431 90" « One whose connection is with “v"4 1” word.
Since this ^*%n"s" `k is not taught in i•*6 }"t$…'o6 /, u|*§ of * }"t$…'o6 / can be studied.
[SK] Q*4q"y*%T- ¬+-*u21s" 1/1 sé/ 1/1 v"431 90" 1/1 R9j" 1/1 -
The 6th case is understood as the one which is connected to “v"4 1”, when the word
to which the 6th case is connected is not known.
¬+Š *u21s5 ¬+*u21s5 (6T) « particularity of connection, to which it is connected.
Q*4q"y*%T5 ¬+*u21s5 3Š"5 " Q*4q"y*%T ¬+*u21s" = sé/ (116B) « one whose
particularity of the connection is not determined, the one to which it is connected to is not
determined.
For example, in 6.1.77 p$9 3&*S «, the 6th case-ending word “p$5” has no ¬+ to any
word in the sūtra. This 6th case is Q*4q"y*%T ¬+*u21s" sé/. When that is the case, this
^*%n"s" `k connects the 6th case to the word “v"4 1”. Now, p$5 6/1 v"4 1 7/1 (in the place of p$)(
makes sense.
94
Qˆ *+5
The next ^*%n"s" `k gives an extra technical meaning to 7th case when the 7th case is
not understood in the senses of the natural language.
This kind of 7th case is called ^"*%n"*s$- â'/, and conventionally called ^%- â'/.
[^*%n"s" `k ']( 1.1.66 y «

T*¼*ü*T *4*oýU 1 ^kuŠ
The $"3y, change, is enjoined in the place immediately before the 7th case-ending
word.
( p*T 0 *4*oýU 1 7/1 ^kuŠ

T*¼4 7/1 y 6/1 «
• T*¼4 ( 7/1 – By the force of p*T, it conveys “any word in 7th case”.
• p*T 0 – This word converts words in sūtra from 2²^% into Q‡^y %. What is pointed is not
the word ”T*¼4”( itself, but any word presented in the 7th case. See ^*%n"s"#$%&' for
(
detail.
• ( say) + ´ (...ed) = that which is said without gap;
*4*oýU 1 7/1 – *4% ((without gap) + *o2 (to
in *T â'/, '"4"*q$%& to T*¼4. (
• y 6/1 – “$"3y'”( is supplied; in ¬+sé/.
^kuŠ
( $"3y' 1/1
[LSK] â'/*4oË241 3/1 *uq/3'"4' 1/1 ( u&"yt%1& 3/1 QZu*‹TŠ 6/1 ^kuŠ ( -
y 6/1 R9j' 1/1
understood as the one which is immediately previous to that 7th case-ending word.
With this ^*%n"s", the 7th case-ending word “Q*S” in 6.1.77 p$9 3&*S «, is understood
as “the $"3y has to be immediately ^ku y to what is told in 7th case”.
( ^%, that which is
To make a u|*§, Q*S 7/1 ^%1 7/1 both in *T â'/, meaning “when QS is
following”, can be said.
95
Even though this sūtra is studied much later in i•*6 }"t$…'o6 /, it can be introduce
y « since they are mirror image to each other. As the
together with 1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
y ) the word
previous sūtra taught that the effect should take place immediately before (^kuŠ
in 7th case, the next sūtra teaches that the effect should take place immediately after (^%Š,
ƒ§%Š) the word in 5th case.
There is no conventional name for this ^"*%n"*s$^Í'/. Yet, we shall call it ^ku^y Í'/ by
its context.
[^*%n"s" `k ']( 1.1.67 6

T¼"*oe§%Š « ~ *4*oýU"T (
( p*T 0 ƒ§%Š 6/1 « ~ *4*oýU"T 5/1

T¼"T 5/1 (

• (
• (
gap; after *un*´*u^*%&"', in *oÏ3901 ^Í'/, '"4"*q$%& to T¼"T; (
• ƒ§%Š 6/1 – “$"3y'”( is supplied. ¬+sé/
( $"3y' 1/1
[LSK] ^Í'/*4oË241 3/1 *uq/3'"4' 1/1 ( u&"yt%1& 3/1 QZu*‹TŠ 6/1 ^%Š 6/1 (= ƒ§%Š 6/1) !13' (
1/1
«
The diagram below can give a general idea of these three ^*%n"s" `k s.
5 6 7
96
Qˆ *+5
( the substitute in the

With the help of the two ^*%n"s" `k s, we understood that 3& is
place of p$ ( which is immediately before QS. ( In our example, the , which is p$,( after q is
(
immediately before ƒ, which is QS. (
6 + ƒ^"Š
q/
( +
q6 3 ( ƒ^"Š 6.1.77 p$9 3&*S
Now, we have another situation. There are four letters in 3&. ( To decide which letter
should be the substitute, we need another ^*%n"s" `k .
3, ( u, ( %,( i (
This ^*%n"s" `k teaches when there are many possible wo12s, the most similar to the
v"*446( should be the wo12.
[^*%n"s" `k ']( 1.1.50 v"41›t%T'5 «

(
When there are many possible wo12s, the most similar to the v"*44 should be
chosen.
v"4 1 7/1 Qt%T'5 1/1 «

2 words in `k , no Q4u|6 *§ is required.
• v"4 1 7/1 – v"4 here means # ¹, possibility of multiple wo12s; in *T â'/.
• (
Qt%T'5 1/1 – Qt% here means ”2, similar. To that, the superlative T*}T suffix T'^ is
added. The meaning is “the closest”. In which way it is the closest is discussed below.
6
(
v"4' QŠ (
Q*Æ p*T v"4/ « ('T^6 -Q‡Ë p*4•-#e35), the one who has/occupies the v"4 for the wo12 is
called v"*44. ( In the example diagram, is occupying the v"4 for the wo12. Thus that is called
v"4/.
97
(
[LSK] # ¹1 7/1 (= v"4 1 7/1) *T 7/1 ”2T'5 1/1 (= Qt%T'5 1/1) wo125 1/1 Š"T III/1 «
When there are multiple possibilities (in wo12), the most similar one (to the v"*44)(
should be the wo12.
By which aspect will the similarity be measured? There are four aspects for
measuring the similarity:
4 types of similarity (wt3y'):(
1. v"4T5 (by the point of articulation)
2. Q‡Ty 5 (by the meaning)
6
3. 0&T5 (by #eÉ, first examine w8t%, then R"ã)
4. #'"&T5 (by '"`", the length of the sound)
For the details and examples, see ^*%n"s"#$%&'. (
(
Among the multiple possibilities of wo12s, the most similar one to the v"*44 in
terms of one of these four aspects is chosen to be wo12.
The v"4 should be examined first. If v"4 does not determine which wo12 to use,
any of the other three criteria can be utilised, without any priority amongst them.
( 3, ( whose
6 + ƒ^"Š, since the v"4 of is T" , the closest among 3& is
In the case of q/
(
v"4 is also T" . Thus, in the place of , the substitute 3 has come.
[LSK] “ q6 3( ƒ^"Š”
( p*T 0 —"T1 7/1 -
The *T â'/ indicates that one $"3y is done, and the next is yet to come. The –^* *}
is “to be continued…”
98
Qˆ *+5
The next *u*q `k gives a consonant sandhi rule, which gives the doubling of a sound.
[*u*q `k ']( 8.4.47 Q4*S S « ~ QS5 3%5 Ò1 u"

(
There is an optional duplication (*Ò•) of 3% (when QS precedes (
and Q4S (anything
other than QS)( follows.
5 6 7
QS ( 3% ( Q4S (
1
Ò1
Q4*S 7/1 S 0 « ~ QS5 5/1 3%5 6/1 Ò1 1/2 u" 0

• ( Q4S (NT)
Q4*S 7/1 – 4 QS p*T ( T*¼4 «( in ^% â'/ (1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
y «)
This 4È is (
( # ·-*4s1q, thus Q4S here brings ‹i ( and Qu "4. The details follow.
• S 0 – connects with the previous sūtra.
• QS5 5/1 – #e"‹"%5 QS; ( in ^ku^y Í'/ (1.1.67 T¼"*oe§%Š
6 «)
• 3%5 6/1 – #e"‹"%5 3%;( in v"431 90" sé/ (1.1.49 sé/ v"431 90" «)
• ( this is wo125. “Becoming two” is commonly called “*Ò•'”.(
Ò1 1/2 – #"*T^*o$' *Ò;
• u" 0 – This rule is optional. Thus there will be two forms.
[LSK] QS5 5/1 ^%Š 6/1 3%5 6/1 (v"4 1 7/1) Ò1 1/2 u" 0 Æ5 II/2 4 0 T 6 0 Q*S 7/1«
There is optional doubling in the place of 3%,( which is after QS, ( but is not followed
by QS. (
6
“^%Š” came from the ^*%n"s" 1.1.67 T¼"*oe§%Š «.
“v"4 1” is understood by the ^*%n"s" 1.1.49 sé/ v"431 90" «.
6 + ƒ^"Š
q/
( +
q6 3 ( ƒ^"Š 6.1.77 p$9 3&*S « ~ *‹T"3"' (
The q, ( which is 3%,( is preceded by ƒ, which is QS, ( is followed by 3, ( which is Q4S (
( 3( +
6q q ( ƒ^"Š 8.4.47 Q4*S S « ~ QS5 3%5 Ò1 u"
[LSK] p*T 0 q$"%Š 6/1 *Ò•14 3/1 “ q6 q( 3( ƒ^"Š”
( p*T 0 —"T1 7/1-
By doubling the q, ( when “ q6 q( 3( ƒ^"Š”
( is produced…
99
2 types of negation (#*Ts1q5)
( a negative particle. It has two aspects as discussed below:

4È is
1.^3o6y " -#*Ts1q5 2. # ·-#*Ts1q5
What it does ”Ïœ"‹/ *4s1q$| T (

It brings other members in the It just negates.
same —"*T.
What it mainly *uq15 #"q"Õ' ( #*Ts1qŠ #"q"Õ' (

conveys It mainly conveys *u*q, a positive It mainly conveys *4s1q, a negation.
injunction.
With what it ƒ§%ô14 ‹ 4È ( *¨33" ‹ 4È (

comes It is compounded in '" . It is used with verb.
Example (
QÚ"Û&' w43 « Ú"Û& 4 w43 «
“Bring somebody who is not a “Do not bring Ú"Û&.”
Ú"Û&.” (You are supposed to (As long as you do not bring Ú"Û&,
bring a “*`3 etc., those who are in you can do anything.)
the same —"*T as a Ú"Û&.)
( 7s-
As seen in the chart, the negative particle of 4È-T° 6 '" usually conveys ^3o6y " -
#*Ts1q. However, “Q4*S” of 8.4.47 Q4*S S« conveys # ·-#*Ts1q.
The reason for Q4*S to be taken as # ·-#*Ts1q is to bring Qu "4, the absence of letter,
usually found at the end of a sentence.
If Q4*S was ^3o6y " -#*Ts1q, being ”Ïœ"‹/, it would bring ‹i,( the —"T/3 (that which has
the same —"*T) to QS. ( And if this was the case, it could have been said “‹*i”, instead of
“Q4*S”, which is one '"`" longer.
(
By taking Q4*S as # ·-#*Ts1q, anything other than QS can be there after the 3%.( Even
though Qu "4 is not —"T/3 to QS, ( it can still be the *4*'§ for 8.4.47 Q4*S S«. Thus we can
(
have the forms such as “u"4”( and “u"$”.
100
Qˆ *+5
( sandhi rule.
The next *u*q `k gives —2-•
[*u*q `k ']( 8.4.53 (

ði" —2 ð*2 «
( the place of ði ( when it is followed by ð2. (
There is substitute —2 in
6 7
ði ( ð2 (
1
—2 (
( —2 1/1
ði"' 6/3 ( ð*2 7/1 «

• (
ði"' 6/3 – #e"‹"%5 ði,( all the consonants except semi vowel and nasal; in v"4 1390" sé/
(1.1.49 sé/ v"431 90" «). To show —"*T (a class), plural number can be optionally used for
(
the singular, by 1.2.58 —"e"¤"3"'1$*¼4 RºuS4'ÕT%Š"' «(
• (
—2 1/1 – #e"‹"%5 —2, ( the 3rd of the classes; this is wo125.
• ð*2 7/1 – #e"‹"%5 ð2, ( the 3rd and 4th of the classes; in ^% â'/ (1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
y «)
( «
[LSK] ÐU' 1/1
Clear. No explanation is required. All the words required to make u|*§ are presented
( v"4 1 7/1 —2 1/1
in the sūtra. With ^*%n"s", it will be like: ði"' 6/3 ( Š"T III/1
( ð*2 7/1 ^%1 7/1 «
[LSK] p*T 0 ^kuqy $"%Š 6/1 o$"%5 1/1 -
( replaced by o (.
Thus the first q is
6 + ƒ^"Š
q/
( +
q6 3 ( ƒ^"Š 6.1.77 p$9 3&*S«
( 3( +
6q q ( ƒ^"Š 8.4.47 Q4*S S«
The q, ( which is ði,( is followed by q, ( which is ð2. (
6 o ( q 3( +
( ƒ^"Š (
8.4.53 ði" —2 ð*2«, with the help of 1.1.50 v"4›1 t%T'5«.
Now, the first ô is 6 o ( q 3( , ( and the next ô is ƒ^"Š.
The next sūtra elides the last letter of a ô which ends with conjunct consonants.
( the letter to be elided.
However, a u"*Ty$ for that sūtra negates its application when 3& is
101
The next *u*q `k gives i9^ for a conjunct consonant-ending word.
[*u*q `k ']( 8.2.23 390"tŠ i9^5 « ~ ôŠ

There is substitute i9^ in the place of the last Qi ( of a ô which ends with conjunct
consonants.
390"tŠ 6/1 i9^5 1/1 « ~ ôŠ 6/1

• 390"tŠ 6/1 – 3905 Qt5 3Š TT ( 390"t' (116B),
( TŠ«; this is an adjective to ôŠ; in
¬+sé/ connected to the words added by the next ^*%n"s" `k . 3905 is a !" given to a
group of conjunct consonants defined as 1.1.7 ‹i9›4t%"5 3905 «
• i9^5 1/1 – This is wo125. i9^5 is a !" defined as 1.1.69 Qo2y4 i9^5 «
• ôŠ 6/1 – From Q*q$"% `k 8.1.16 ôŠ «; this sūtra goes down as Q4u|6 *§ until 8.3.54;
( a !" defined as 1.4.14 *â™t
ô' is 6 ( in ¬+sé/ connected to the words added
ô' «;
by the next ^*%n"s" `k
( 3T 1/1
[LSK] 390"t' 1/1 ( ô' 1/1
( TotŠ 6/1 i9^5 1/1 Š"T III/1
( «
There is substitute i9^ in the place of the end of the ô which ends with 390. Note
that Tot in the u|*§ is 6T.
If we try to understand the 6th case as v"431 90" sé/, it will mean “ 390"tŠ ôŠ v"4 1
i9^5”, thus the whole ô gets elided. But this is not pU, a desired end to be achieved. We
need a ^*%n"s" `k to arrive at the correct interpretation.
6
ô 390
1
i9^ ?
Note: Since there is a ^*%n"s" `k 1.1.72 314 *u*qÆotŠ «, which can supply the word “Qt”
between two words in '"4"*q$%&, Pā ini could have said “ 390Š” to save some letters.
He however said “ 390"tŠ” for the sake of clarity and easy comprehension on the part of
students.
102
Qˆ *+5
The next ^*%n"s" `k tells that the last letter of what is presented by 6th case should be
replaced by the substitute.
[^*%n"s" `k ']( 1.1.52 Qi9›•Š «

Qi5 6/1 Q•Š 6/1 «
• Qi5 6/1 – #e"‹"%5 Qi ( any letter; in v"431 90" sé/ (1.1.49 sé/ v"431 90" «)
• Q•Š 6/1 – Qt1 nu5 Q•5« that which exists at the end is Q•5, the last; adjective to
Qi5.
(
[LSK] sé/*4*oýUŠ 6/1 Q•Š 6/1 Qi5 6/1 wo125 1/1 Š"T III/1 «7
The substitute should be in the place of the last letter of what is presented in 6th case.
For example, in 8.2.23 390"tŠ i9^5 « ~ ôŠ, the substitute i9^ should be in the
place of the last letter of what is presented in 6th case, which are 390"tŠ and ôŠ.
In the case of o6 (q3( +
( ƒ^"Š, “ o6 (q3( ”( is the meaning of 390"t ô. Since it contains
more than one letter, we take recourse to 1.1.52 Qi9›•Š «. Then the whole u|*§ will
( «”
become “ 390"tŠ 6/1 ôŠ 6/1 Q•Š 6/1 Qi5 6/1 v"4 1 7/1 i9^5 1/1 wo125 1/1 Š"T III/1
Notice that “ 390"tŠ 6/1 ôŠ 6/1” are '"4"*q$%& and in ¬+sé/ to “Q•Š 6/1 Qi5
6/1
”, which are also '"4"*q$%& and connected to “v"4 1 7/1”.
When this is the case, the last letter of “ o6 (q3( ”( is “3”.( Thus the situation, in which
the substitution i9^ should be in the place of 3, ( is possible. But it is going to be negated by
the next u"*Ty$.
[LSK] p*T 0 3i9^1 7/1 #"â1 7/1 –
Again, *T â'/ suggests the next step.
7
In the Gita Press book, “sé/*4*oýU5” is written, but the 6th case-ending form fits better. Other
commentary books are uniformly using the 6th case-ending form.
103
The next u"*Ty$ is regarding 8.2.23 390"tŠ i9^5«. This u"*Ty$ negates the application
of this sūtra when what is going to be elided is 3&. (
(u"*Ty$')( 3&5 #*Ts1q9 u"Î5 «

3&5 6/1 #*Ts1q5 1/1 u"Î5 1/1 «
• 3&5 6/1 – #e"‹"%5 3&; ( in ¬+1 sé/ to #*Ts1q5.
• #*Ts1q5 1/1 – Prohibition (of elision of 3&).(
• ( say) + Á3T ($'y
u"Î5 1/1 – uS (to ( *& + necessity); that which has to be said.
This should be understood as “3&5 6/1 390"ti9^-#*Ts1q5 1/1 u"Î5 1/1”. Regarding
(
8.2.23 390"tŠ i9^5«, the prohibition of elision of 3& should have been told (but was not
told by `k $"%).
( the end of “ o6 (q3( ”( does not happen.
Thus, the i9^ of the 3 at
[LSK] }6 3( ^6 "Š 1/1 «

The whole –^* *} is as follows:
6 + ƒ^"Š
q/
( +
q6 3 ( ƒ^"Š 6.1.77 p$9 3&*S« ~ *‹T"3"' (
6th case is understood by 1.1.49 sé/ v"431 90" «
7th case if understood by 1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
y «
wo12 is understood by 1.1.50 v"4›1 t%T'5«
( 3( +
6q q ( ƒ^"Š 8.4.47 Q4*S S« ~ QS5 3%5 Ò1 u"
6 o ( q 3( +
( ƒ^"Š (
8.4.53 ði" —2 ð*2«
3-i9^ is #"â by 8.2.23 390"tŠ i9^5 «
The letter for operation is determined by 1.1.52 Qi9›•Š «
3-i9^ is negated by (u"*Ty$')( 3&5 #*Ts1q9 u"Î5 «
}6 3( ^6 "Š u&y ?1i4 1 (combining the letters together)
*Ò•-Qn"u-^“1 (in the case of absence of *Ò• by 8.4.47 Q4*S S«, since it is optional)
j 6
6 ^"Š After applying 6.1.77 p$9 3&*S«, nothing more happens.
104
Qˆ *+5
How to read *»å"&/ which describes –^* *}

Each *T â'/-ending word indicates one $"3y is done. Thus, after each *T â'/-
ending word, put a comma and break the line.
This way of presenting –^* *} in Sanskrit is very useful to see the flow of discussion.
Students can try to adopt this style as a template to do –^* *} orally.
[LSK *»å&/ ]
6 + ƒ^"Š5” p*T *vT1, (when this is the state,)
“ q/
y ” “v"4›1 t%T'5” p*T `k Ò3 ‹$"%1& (with the help of these two sūtras,) “p$9
“T*¼*ü*T *4*oýU 1 ^kuŠ
3&*S” p*T `k & (
1 q$"%-(ƒ§%u*Ty)- $"%Š 3$"%1, (when 3 for ( gained,)
existing after q is
“Q4*S S” pe441 o$"%Š *Ò•1,
(
“ði" —2 ð*2” p*T ^kuqy $"%Š o$"%1,
(
“Qo2y4 i9^5” p*T i9^ !"3"' Qu0T"3"' , ( (when !" is gained,)
“Qi9›•Š” pe1T„‹$"%1& “ 390"tŠ i9^5” pe441 3$"%Š i9^1 #"â1, (when i9^ is possible,)
“3&5 #*Ts1q9 u"Î5” p*T T*üs1q,1 (when there is the prohibition of that,)
u&y ?1i4 1 S $| T1, (when putting the letters together is done,)
6 ^"Š5”
“ ‰ 6 p*T * }' «( (the correct form is achieved.)
6
6 ^"Š5”
[LSK *»å&/ ] *Ò•-Qn"u1 “ j p*T (* }')( «
6
6 ^"Š5”.
When there is absence of *Ò• (by the optional rule, Q4*S S«), the form is “ j
[LSK *»å&/ ] QŠ “^*Á¸T5• w%"q4/35” pe‡5y «
The meaning of this is “object to be worshipped by the learned people”.
[LSK] '@*%5 « q"`25« i"$| *T-

By the nature of the sūtra 6.1.77 p$9 3&*S«, we can assume there must be four types
of examples, namely, pu&y + QS, ( ƒu&y + QS, ( xu&y + QS, ( and {u&y + QS. (
The examples shown in LSK are; 'q 6 + Q*%5, q"T| + Q25, { + w$| *T5.
105
Homework:
1. Following the style below, write the line-by-line –^* *} of the four examples with
optional forms as well.
6 + ƒ^"Š
q/
( +
q6 3 ( ƒ^"Š 6.1.77 p$9 3&*S«
6th case is understood by 1.1.49 sé/ v"431 90" «
7th case if understood by 1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
y «
wo12 is understood by 1.1.50 v"4›1 t%T'5«
( 3( +
6q q ( ƒ^"Š 8.4.47 Q4*S S«
6 o ( q 3( +
( ƒ^"Š (
8.4.53 ði" —2 ð*2«
3-i9^ is #"â by 8.2.23 390"tŠ i9^5 «
The letter for operation is determined by 1.1.52 Qi9›•Š «
3-i9^ is negated by (u"*Ty$')( 3&5 #*Ts1q9 u"Î5 «
}6 3( ^6 "Š u&y ?1i4 1 (combining the letters together)
2. Describe orally these –^* *} in the form of *»å&/. The template is given below.
“ ” p*T *vT1,
6/1 S7/1
“ ”“ ” p*T `k Ò3 ‹$"%1& “ ” p*T `k &
1 ,
6/1 S7/1
“ ” pe441 ,
6/1 S7/1
“ ” p*T ,
S7/1 S7/1
“ ” p*T ,
6/1 S7/1 S7/1
“ ” pe1T„‹$"%1& “ ” pe441 ,
S7/1
“ ” p*T ,
u&y ?1i4 1 S $| T1,
“ ” p*T * }' «(
106
Qˆ *+5
Qˆ *+5 2) Q3u"3"u- *+5 “‹%31”
The next *u*q `k gives the “Q3u"3"u- *+ rule”, as in the diagram below.
6 7
‘S ( QS (
1
Q3, ( Qu, (
w3, ( wu (
[*u*q `k ']( 6.1.78 ‘S9›3u"3"u5 « ~ Q*S *‹T"3"' (

(
‘S5 6/1 Q3u"3"u5 1/3 « ~ Q*S 7/1 *‹T"3"' 7/1
• ‘S5 6/1 – #e"‹"%5 ‘S; ( in v"431 90" sé/ (1.1.49 sé/ v"431 90").
With the ^*%n"s", it results in “‘S5 6/1 v"4 1 7/1”, “in the place of ‘S”(
• ( Qu S
Q3u"3"u5 1/3 – Q3 S ( wu S;
( w3 S ( pT%1T%Ò) '" 5. This is wo125.
• ( ^% â'/ (1.1.66 T*¼*ü*T *4oËU 1 ^kuŠ

Q*S 7/1 – From 6.1.77 p$9 3&*S «; #e"‹"%5 QS in y )
( following”
With the ^*%n"s", it results in “Q*S 7/1 ^%1 7/1”, “when QS is
• (
*‹T"3"' 7/1 – From Q*q$"% `k 6.1.72 *‹T"3"' ;( *‹T" is a !" defined as 1.4.109 ^%5
*ü$sy5 *‹T"; in *us3 â'/; “in the topic of *‹T"”
( *us31 7/1 «
Our own u|*§ will be – ‘S5 6/1 v"4 1 7/1 Q3u"3"u5 1/3 Š56 III/3 Q*S 7/1 ^%1 7/1 *‹T"3"' 7/1
(
In the topic of *‹T", Q3u"3"u are the substitutes in the place of ‘S, ( when QS is
(
following.
( Qu 1/1
[LSK] ‘S5 6/1 ¨'"T 0( “Q3 1/1 ( wu 1/1
( w3 1/1 ( ” ‘T1 1/3 (= Q3u"3"u5 1/3) Š56 III/3 Q*S 7/1 -
(
In i•*6 }"t$…'o6 /u|*§, the word “¨'"T (respectively)” is inserted. This word is the
essence of the next ^*%n"s" `k 1.3.10 3‡" û'4o16 25 '"4"' «(
107
See the sūtras in QU"j"3/ `k ^"l.

6.1.72 (
*‹T"3"' V7/1 «
6.1.77 (
p$5 6/1 3& 1/3 Q*S 7/1 « (
*‹T"3"' V7/1
6.1.78 ‘S5 6/1 Q3u"3"u5 1/3 « Q*S 7/1 (
*‹T"3"' V7/1
Observe how the Q4u|6 *§ (bringing down of words from previous sentences) works.
This system is observed in the natural language as well. A word in the previous sentence
is overwritten by another word in the following sentence when the case ending is the same.
And when a particular case ending word in the previous sentence is not found in the
following sentence, that word is brought down to the following sentence as Q4u|6 *§.
Now, we have a situation. There are four v"*44s( (those which are occupying the
v"4, the place for substitution) and four wo12s. We need a ^*%n"s" to decide which v"*44 (
should be replaced by which wo12. 1.1.50 v"4›1 t%'5 « will not do because the v"4 of both ‘
and ¾ is $ÁlT" , and the v"4 of both , and ç is $Ál9é'. ( We cannot decide by the
similarity of v"4 or even R"ã#3É, Q‡,y or '"`". There is an QZuv" (situation in which there
is no definiteness). We need another ^*%n"s" `k for the correct interpretation of ‘S9›3u"3"u5 «.
The gist of this sūtra in one word is “respectively”. As seen in the diagram below,
when a rule involves two sets of things which are the same in number, those two things
are connected respectively in the order of appearance.
‘ , ¾ ç
Q3 ( Qu ( w3 ( wu (
More detailed explanation of this sūtra is “when a rule has two groups which have
the same number of members, those members should be matched by the number in order
of appearance”.
108
Qˆ *+5
[^*%n"s" `k ']( 1.3.10 3‡" û'4o16 25 '"4"' «(

( «
3‡" û' 0( Q4o16 25 1/1 '"4"' 6/3
3 words in the `k ; no word as Q4u|6 *§
• 3‡" û' 0( – This is QZ3/n"u '" . *uœ‹u"À is “ û"' f/2/1
( (number) Q4*T¨Å (not
exceeding, not transgressing) p*T 3‡" û'”.( This word is used as adverb.
• Q4o16 25 1/1 – Any rules told in sūtras.
• (
'"4"' 6/3 – of two groups which have the same number of members
( ) *u*q5 1/1 (= Q4o16 25 1/1) 3‡" û' 0( Š"T III/1

[LSK] ' ¬+/ 1/1 (= '"4"' 6/3 ( «
' ¬+/ is adjective to *u*q5. ' ¬+/ means “the one which has the connection
between the same number.” '"4- û"4" ¬+5 ' ¬+5 « The connection ( ¬+) between
the same ( '"4) numbers ( û") is ' ¬+. ' ¬+5 QŠ Q*Æ p*T ' ¬+/ « ' ¬+ + p*4•
('•‡ y-T*}T-#e3, one which has …) = ' ¬*+4, ( ' ¬*+4 in
( masculine 1/1 is ' ¬+/.
In this case, *u*q = Q4o16 2 is the sūtra “‘S9›3u"3"u5«”. There is a connection between
(
‘S and Q3u"3"u. ( Both of them have the same number of members, four.
When this is the case, the rule (Q4o16 2) should be (Š"T)( 3‡" û', ( according to the
number in the order of appearance.
u|*§ continues to show examples.

[LSK] ‹%31 4/1 « *uŽu1 4/1 « 4"3$5 1/1 « ^"u$5 1/1 «
Since the sūtra consists of 4 combinations, we have to see 4 types of examples.
Example 1: ‹%31
‹*% + ™1 Declension of #"*T^*o$ “‹*%” in 4th case singular
‹*% + ‘ 6
Q4R+i9^5 (
(elision of pT letters)
‹%1 + ‘ 6 Q¹Š (We have not studied this sūtra yet.)
7.3.111 •1*™y *T « ~ 0&5
( ‘
‹%3 + 6.1.78 ‘S9›3u"3"u5 « ~ Q*S *‹T"3"' (
With the help of 1.3.10 3‡" û'4o16 25 '"4"' «(
‹%31 (
u&y ?1i4' (putting all the letters together)
109
Example 2: *uŽu1
*uŽ 6 + ™1 Declension of #"*T^*o$ “*uŽ”6 in 4th case singular
*uŽ 6 + ‘ 6
Q4R+i9^5 (
*uŽ9 + ‘ 6 Q¹Š (We have not studied this sūtra yet.)
7.3.111 •1*™y *T « ~ 0&5
( ‘
*uŽu + 6.1.78 ‘S9›3u"3"u5 « ~ Q*S *‹T"3"' (
*uŽu1 (
Example 3: 4"3$5
(
&/È #"^&1 (1U) to lead
6 (
4/ + Áui (
3.1.133 Áu¢6 T|S… « ($Ty*% $| T-#e35; see volume 3)
4/ + u 6 6
Q4R+i9^5 (
4/ + Q$ 7.1.1 3u6 9%4"$… « (u 6 is replaced by Q$; see volume 3)
4 • + Q$ (
7.2.115 QS9 *Ø&*T « ~ u|*}5 Q¹Š (QS followed (
by *&T takes u|*}: See volume 3)
With the help of 1.1.50 v"4›1 t%T'5 «
(
4"3 Q$ 6.1.78 ‘S9›3u"3"u5 « ~ Q*S *‹T"3"' (
4"3$ (
Example 4: ^"u$5
(
^kÈ ^"u4 1 (9U) to purify
6 (
^k + Áui (
3.1.133 Áu¢6 T|S… « ($Ty*% $| T-#e35; see volume 3)
^k + u 6 6
Q4R+i9^5 (
^k + Q$ 7.1.1 3u6 9%4"$… « (u 6 is replaced by Q$; see volume 3)
^… + Q$ (
7.2.115 QS9 *Ø&*T « ~ u|*}5 Q¹Š (QS followed (
by *&T takes u|*}: See volume 3)
With the help of 1.1.50 v"4›1 t%T'5 «
(
^"u +Q$ 6.1.78 ‘S9›3u"3"u5 « ~ Q*S *‹T"3"' (
^"u$ (
110
Qˆ *+5
Before studying the next sūtra, we shall study two ^*%n"s" `k s, which are not in
i•*6 }"t$…'o6 /, but very much essential for understanding the next sūtra and many others.
The next ^*%n"s" `k tells that when two words are in '"4"*q$%&, and in *u21s&
(adjective) / *u21³ (qualified noun) relationship, the adjective is understood both as “that
which ends (= Qt) with the adjective”, and also “the adjective itself”.
This operation is conventionally known as “Tot*u*q5”.
[^*%n"s" `k ']( 1.1.72 314 *u*qÆotŠ « ~ €Š –^Š

When *u*q is told with adjective, that adjective is represents for that which ends
with that adjective, and also for that adjective itself.
314 3/1 *u*q5 1/1 TotŠ 6/1 « ~ €Š 6/1 –^Š 6/1

• 314 3/1 – By which (adjective); it is understood as “314 *u21s&14”; in $%&1 T|T/3".
• ( *u21s4')( Qt1 3Š 5 Tot5 (176B) = 2² 'o6 "35 (group of letters), TŠ «;
TotŠ 6/1 – TT (=
“for the group of letters which ends with that adjective”; in ¬+sé/ connected to “ !"”
making a sentence “the adjective is a !" for the group of letters which ends with that
adjective”.
• €Š 6/1 – #"*T^*o$ € (its own) indicating *u21s&; in ¬+sé/, connected to “–^Š”
• –^Š 6/1 – Of the form; in ¬+sé/, also connected to “ !"”; “The adjective is also a
!" for its own form”
( TotŠ 6/1 !" 1/1 Š"T III/1

[SK] *u21s&' 1/1 ( €Š 6/1 S 0 7^Š 6/1 «
Adjective is a !" for a group of words ending with that adjective, and also for that
adjective itself.
( « ~ q"T95 5/1 #e35 1/1, ‘5 5/1 (#"*T^*o$ is p)

For example, in the case of 3.3.56 ‘5 5/1 QS 1/1
is an adjective to q"T95 5/1. By this ^*%n"s" `6 1.1.72 314 *u*qÆotŠ «, the *u21s& “p” is
understood as a !" for “p-ending q"T”6 , and also “p, which is q"T”6 .
111
The next u"*Ty$ is Qû"o to the Tot*u*q. This operation is known as To"*o*u*q.
When the three conditions below are satisfied, the adjective is understood as “that
which begins (= w*o) with the letter”.
Three conditions for To"*o*u*q:
1. Two words are in '"4"*q$%&, in a *u21s& (adjective)/*u21³ (qualified noun)
relationship.
2. They are in ^% â'/.
(
3. The *u21s& (adjective) refers to any letter (Qi-œ‹&).
(u"*Ty$')( 3*¼*µ*qÆo"o"u¢œ‹&1 «
When Qi ( is mentioned as an adjective, and told in 7th case, then it should be
understood as that which begins with that letter.
( *u*q5 1/1 To"o… 7/1 Qi-œ‹&1
3*¼4 7/1 ( 7/1
«
• (
3*¼4 7/1 – “p*T â'/*4*oýU”1 is also understood; in *T â'/; '"4"*q$%& to Q¢œ‹&1;
• *u*q5 1/1 –An injunction, rule, told by sūtra.
• To"o… 7/1 – TT (= ( in
( Qi)( wo… 3Š 5 To"*o5 (176B) = 2² 'o6 "35 (group of letters), T*¼4 «;
^% â'/. “p*T u1*oTZ'”( should be understood.

• (
Q¢œ‹&1 7/1 – Qi ( (any letter) œ‹&' (mentioned) (
Q¢œ‹&' (KT) ( in *T â'/.
T*¼4 «;
( Q¢œ‹&1 7/1 *T 7/1 *u*q5 1/1 $"3y' 1/1

[Medhā] 3*¼4 7/1 ( nu*T III/1, “To"o… 7/1 ^%1 7/1” p*T 0 u1*oTZ' 1/1
( «
If an operation takes place when a condition follows, and the condition is

mentioned as a letter, then it should be understood as “that which begins with that letter”.
An example for this is seen in the next sūtra.
112
Qˆ *+5
Coming back to i•*6 }"t$…'o6 /, the next *u*q `k is similar to ‘S9›3u"3"u5, with more
limited scope as shown below.
6 7
,, ç 3( #e3
1
Qu, ( wu (
[*u*q `k ']( 6.1.79 (

u"t9 *3 #e31 « ~ *‹T"3"' (‘S5)
( (,o…T95 6/2, by u"*Ty$ or Q4'"4)
u"t5 1/1 *3 7/1 #e31 7/1 « ~ *‹T"3"' 7/1 6

• (
u"t5 1/1 – u5 Qt1 3Š 5 u"t5 (176B) « u$"%"T Q$"%5 ( the
ƒí"%&"‡5y « That which has u at
(
end is u"t. Among the wo12s in the previous sūtra, u"ts are Qu and wu. ( This is wo125.
• ( ^% â'/.
*3 7/1 – The letter 3 in
• #e31 7/1 – in ^% â'/.
Since these two words are in '"4"*q$%& (condition 1), and they are both in ^% â'/
(
(condition 2), and one of them is mentioning a letter (Qi-œ‹&) (condition 3), To"*o*u*q is
applied. Thus “3$"%"o… #e31” or “3$"%"*o#e31” is understood.
• ‘S5 6/1 (,o…T95 6/2) – #e"‹"%5 ‘S; ( in v"431 90" sé/ (1.1.49 sé/ v"431 90")
(
Because wo12s are only Qu and (
wu, ( the v"*44 are understood as corresponding ,
( added between , and ç, and ç and *un*´ , in
and ç only. For clear presentation, T is (
( çT =
order to prevent vowel sandhi. In ,T + ( ,o…T (ID),
( ( ^ku^y o is ô"t, thus
the T of
becomes o ( by 8.2.39 ði" —29›t1 «.
• (
*‹T"3"' 7/1 – This has not much importance since this is between #$| *T and #e3.
Our own u|*§ will be – ,o…T95 v"4 1 Qu, ( wu p*T
( u"t5 wo125 Š"T 3$"%"o…
( #e31 ^%1«
( ”( ‘T… 1/2 Æ5 III/2 «

[LSK] 3$"%"o… 7/1 #e31 7/1 ^%1 7/1 ,o…T95 6/2 “Qu wu
(
Qu and wu, ( these are the substitutes in the place of , and ç, respectively, when 3-(
beginning suffix follows.
113
“*3” of the sūtra is understood as “3$"%"o…” by applying To"*o*u*q. “^%1” by

T*¼*ü*T*4*oýU 1 ^kuŠ (
y «. “,o…T95” is the modification of the Q4u|6 *§ ‘S5. “Qu wu (
‘T…” is the
explanation for “u"t5” in the sūtra.
Now we need to see two ƒo"‹%&s:

1. , followed by 3$"%"*o#e3
2. ç followed by 3$"%"*o#e3.
( «
[LSK] 0Z' 1/1
095 *u$"%5 (product of cow) 0Z oj"*o (such as curd)
( 3T ( 4.3.160 09^3 93yT «( ~ TŠ *u$"%5 T*}T"5 #e35
09 (cow) + ™ +
09 + 3 6
Q4R+i9^5, 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
( +
0 Qu ( 3 6.1.79 u"t9 *3 #e31 «
0Z
( «
[LSK] 4"Z' 1/1
(
4"u" T"3y' (that (
which is crossed by boat) 4"Z —i' (water)
4… (boat) + »" + 3T ( 6 #"Auj"4"Å ' *'T *'T1s 6 «
6 "8Æ"3yTÞ
4.4.91 4…u39q'y*us'ki'ki /T"Ti
~ 3T (
4… + 3 6
Q4R+i9^5, 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
( wu +
4+ ( 3 6.1.79 u"t9 *3 #e31 «
4"Z
114
Qˆ *+5
Now one u"*Ty$ on 6.1.79 u"t9 *3 #e31 « is taught in i•*6 }"t$…'o6 /. This u"*Ty$ is
related to another u"*Ty$ which is not seen in i•*6 }"t$…'o6 /.
They are to address the form “0Zk*T”, which is made of 09 and 3k*T. Since 3k*T is
#"*T^*o$, not #e3, u"*Ty$$"% composed these u"*Ty$s.
(u"*Ty$')( 093T
k y … ¶¡Š^6 û"4' «(
In the topic of u"t-wo12, “09” followed by a #"*T^*o$ “3k*T” in the Vedas should be
told additionally (ƒ^ û"4').(
095 6/1 3kT… 7/1 ¶¡* 7/1

( «
ƒ^ û"4' 1/1
Thus the word “0Zk*T” word is seen in the Vedas. The meaning is “09#S%nk*'5”,
pasture.
The next u"*Ty$ is for the form 0Zk*T in non-Vedic literature.
(u"*Ty$')( Qè^*%'"&1 S «
Even in i9$ (non-Vedic literature), the word “0Zk*T” word is used, but only in the
sense of the measurement of distance.
Qè^*%'"&1 7/1 S 0 «
• ( (measurement) Qè^*%'"&' (6T);
Qè^*%'"&1 7/1 – Qè45 6/1 (of the road) ^*%'"&' 1/1 ( in
*us3 â'/.
• S 0 – This brings 3k*T and u"t5.
(
[*»å&/ "] 09-2²-,$"%Š 6/1 3k*T-2²1 7/1 ^%1 7/1 Qu-wo1
25 1/1 i9$1 7/1 Q*^ 0 nu*T III/1 '"0y^*%'"&1 7/1
0Å1 7/1 «
(
When the measurement of a road is understood, even in non-Vedic literature, Qu is
the substitute for the , of 09-word when 3k*T-word is following.
[*»å&/ "] 0Zk*T5 1/1 p*T 0 ¨92Ò3Š 6/1 !" 1/1 «
0Zk*T is a name for two ¨92, another unit of measurement of distance.
115
6 *+5 “ƒ^1Œ5”
Qˆ *+5 3) 0&-
6 in the next !" `k .

First we will see the definition of the technical term 0&
[ !" `k ']( 1.1.2 6 «

Qo1™ ( 0&5
6
Q, ‘, and , - These three letters are termed 0&.
( ‘™ ( 1/1 0&5
QT 1/1 6 1/1 «
3 words in the `k . No Q4u|6 *§ is required.

• (
QT 1/1 – This is !/. By 1.1.69 Q&*6 oT ( u&yŠ S"#e35 «, letter Q represents 18 varieties of
(
Q. However, because of the T after Q, the Q in this sūtra represents only 6 varieties of
•€ Q, with the help of ^*%n"s" 1.1.70 T^%ÆB"iŠ«, introduced next.
This Q is called T^%, that from which T is (
( ^%. (T5 1/1 ^%5 1/1 3¼"T 5/1 5 1/1 T^%5 1/1 = Q)
• ‘™ ( 1/1 – #e"‹"%5 ‘™;( indicating ‘ and ,. This is !/. This ‘™ ( represents only 6 varieties
of o/• y ‘ and , each, with the help of ^*%n"s" 1.1.70 T^%ÆB"iŠ«, introduced next.
This ‘™ ( is also called T^%, that which is after T. ( (T"T 5/1
( ^%5 1/1 T^%5 1/1 = ‘™)(
• 6 1/1 – This is !".

0&5
( ‘™ ( 1/1 S 01 0&
[LSK] QT 1/1 (
6 !5 1/1 Š"T III/1 -
6
Short Q and ‘™ ( (‘ and ,) are termed 0&.
( this sūtra, Q and ‘™ ( can only bring their u&ys which are of the
Because of the T in
6
same length as themselves. Thus only •€-Q$"% and o/• y ‘ and , are termed 0&.
This $"i*43' (restriction of time) is taught in the next sūtra.
116
Qˆ *+5
The next sūtra gives $"i*43', restriction of time, or length, to the u&ys, which a
vowel can represent as instructed by the previous sūtra 1.1.69 Q&*6 oT ( u&yŠ S"#e35«.
[ !" `k ']( 1.1.70 T^%ÆB"iŠ « ~ u&yŠ

T^% (that which is followed by T, ( and also that which is after T)( is !" for u&y with
the same length.
T^%5 1/1 TB"iŠ 6/1 « ~ u&yŠ 6/1

• T^%5 1/1 – This is adjective to vowel.
There are two ways of solving this '" .
(
1. T5 1/1 ^%5 1/1 3¼"T 5/1 ( ^% (following)
5 1/1 T^%5 1/1 (115B); that from which T is
Ex. QT, ( wT, ( pT, ( ‘T, ( ¾T, ( ,T, ( etc.
( ^%5 1/1 T^%5 1/1 (5T); that which is after T (
2. T"T 5/1
( ( 0&5”.
Ex. ‘™ ( in the sūtra “QT ‘™ 6
• TB"iŠ 6/1 – TŠ (T^%•14 ƒí"3y'"&Š) $"i5 pu $"i5 3Š 5 (616B, IB) «; TB"i is an

adjective to u&y, in ¬+sé/ to !", together resulting in “TB"iŠ u&yŠ !"”.
(
[LSK] T5 1/1 ^%5 1/1 3¼"T 5/1 ( ^%5 1/1 S 0 (= T^%5 1/1) ƒí"3y'"&- '-$"iŠ 6/1 (TB"iŠ
5 1/1, T"T 5/1
6/1
(
) ‘u 0 !" 1/1 Š"T III/1 «
T^% (that which is followed by T, ( and also that which is after T)( is !" only for that
which has the same length as being uttered.
Note that ‘u-$"% (the word “‘u”) is to indicate restriction.
This sūtra is called $"i*43' to the previous sūtra 1.1.69 Q&*6 oT ( u&yŠ S"#e35 «,
(
teaching that Q& (all the vowels and semivowels) is a !" for all the u&ys.
(
By adding T before or after a vowel, that vowel is called a T^% vowel, and it is a !"
for u&ys which have the same time ('"`") as the time of that T^% vowel.
( a !" for 6 varieties of •€ Q (Q, Q8, Q7 , Q•, Q• 8, Q•7 ).
For example, QT is
117
[Examples]
( T^%, thus it is a !" only for u&y with the same time, which is •€.
1. Q of QT is
Q w wæ Q• w• w•æ
Q8 w8 w8æ Q• 8 w• 8 w• 8æ
Q7 w7 w7æ Q•7 w•7 w•7æ
( ”( is T^%, thus it is a !" only for u&y with the same time, which is o/•.y
2. ‘™ ( of “QT ‘™
- ‘ ‘æ - ‘• ‘•æ - , ,æ - ,• ,æ•
- ‘8 ‘8æ - ‘• ‘•æ - ,C ,C æ - ,• 8 ,æ• 8
- ‘7 ‘7æ - ‘•7 ‘•æ7 - ,7 ,7æ - ,7 • ,7æ•
( a !" only for u&y with the same time, which is o/•.y
3. T is
p æ p• D Dæ
p8 C Cæ p• 8 DC DCæ
p7 7 æ7 D7 D7 Dæ7
Note:
( a vowel”. There are two
“T^%-$%&” means “making (a vowel) T^%” or “attaching T to
types of T^%-$%&.
1. T^%-$%& with 1.1.70 T^%ÆB"iŠ « :
This is what we just studied.
2. T^%-$%& just for clarity of the sound:
As seen in “‘o•T95”, “,o…T95”, etc., the T$"% is used for clear identification of the
vowels (by preventing sandhi taking place between them).
118
Qˆ *+5
[*u*q `k ']( 6.1.87 6 « ~ ‘$5 ^ku^y %395 Q*S *‹T"3"' (

wo ( 0&5
(
Qu&y followed by QS together 6
becomes 0&.
5 7
Qu&y QS (
6
1
6
0&5
( 0&5
wT 5/1 (
6 1/1 « ~ ‘$5 1/1 ^ku^y %395 6/2 Q*S 7/1 *‹T"3"' 7/1

• (
wT 5/1 – #"*T^*o$ is Q, in 5th case (^ku^y Í'/ by 1.1.67 T¼"*oe§%Š«)
6
This Q is Q*uq/3'"45, thus it stands for 18 u&ys as per 1.1.69 Q&*6 o„u&yŠ S"#e35«.
• 6 1/1 – 0&-
0&5 6 !" is told by 1.1.2 Qo1™ ( 0&5«.
6 This is wo125.
• 6
‘$5 1/1 – From an Q*q$"%- `k 6.1.84 ‘$5 ^ku^y %395«. This is qualifying the 0&-wo1
25.
• ^ku^y %395 6/2 – From an Q*q$"%- `k 6.1.84 ‘$5 ^ku^y %395«, in v"431 90" sé/; “in the place of ^ku y
and ^%”
• ( ^% â'/ (1.1.66 T*¼*ü*T *4oËU 1 ^kuŠ
Q*S 7/1 – From 6.1.77 p$9 3&*S; #e"‹"%5 QS in y )
( Q*uq/3'"45, thus it stands for all the u&ys as per 1.1.69 Q&*6 o„u&yŠ S"#e35«.
This QS is
• (
*‹T"3"' 7/1 – From the Q*q$"% `k 6.1.72 *‹T"3"' (
Note: See sūtras under the Q*q$"% of 6.1.84 ‘$5 ^ku^y %395 «. All the wo12s in this section
are ‘$"o12, one substitute for two. Check the 1st case ending words, which are wo12s, in this
section. Starting from 0&5 (
6 in this sūtra, u|*}5 (6.1.88 to 92), ^%–^' (6.1.94 to 100), o/•5y (6.1.101
to 106), ^ku– (
y ^' (6.1.107 (
to 110), and ƒT (6.1.111) are all ‘$"o125.
( (= wT 5/1
[LSK] Qu&"yT 5/1 ( ) Q*S 7/1 ^%1 7/1 ^ku^y %395 6/2 ‘$5 1/1 0&5 (
6 1/1 wo125 1/1 Š"T III/1 «
6 is the one substitute in the place of both ^ku y and ^%. What precedes is Qu&y and
0&
what follows is QS. ( An expression “~u&y” indicates all the u&ys. Thus Qu&y means all the 18
u&ys of Q.
119
Now we will see the application of this sūtra in examples. Taking the words of the
sūtra literally, we can think of 9 combinations as follows. However, the cases #1 and #6 to
#9 are covered by other rules, which are called Qû"o.
1) Qu&y + Qu&y =w (o/•5y ) o•e"*%5 [Qû"o5] 6.1.101 Q$5 u&Ë o/•5y «

2) Qu&y + pu&y =‘ 6
(0&5) ƒ^1Œ5
3) Qu&y + ƒu&y =, 6
(0&5) 0¹9o$' (
4) Qu&y + xu&y 6
= Q% ( (0&5) $| Ž*}y5
5) Qu&y + {u&y = Qi ( (0&5)
6 TuF"%5
6) Qu&y + ‘u&y =¾ (u|*}5) $| Ž•$•' ( [Qû"o5] 6.1.88 u|*}%1*S «
7) Qu&y + ,u&y =ç (u|*}5) 0¹…•5 [Qû"o5] 6.1.88 u|*}%1*S «
8) Qu&y + ¾u&y =¾ • 3y' (
(u|*}5) o1uÂ [Qû"o5] 6.1.88 u|*}%1*S «
9) Qu&y + çu&y =ç (u|*}5) $| Ž…BÁG' ( [Qû"o5] 6.1.88 u|*}%1*S «
When there are two rules applicable in one place at the same time, the rule which
will become useless if it is not applied is called Qû"o.
6 and u|*} rules are applicable at the same time in Q + ‘. If 6.1.87
For example, both 0&
w &5
6 « were to be applied, 6.1.88 u|*}%1*S « would become useless. Therefore 6.1.88 u|*}%1*S «
is called Qû"o5 and 6.1.87 w &5
6 « is called ƒ„0y5.
When there is such a relationship between two sūtras, Qû"o sūtra should be
applied and ƒ„0y sūtra should be dismissed.
^% Q/w p/ ƒ/ù x/9 { ‘ , ¾ ç
QS (
^ku y u&y5 p$ ( ‘S (
Q/w o/qy5 6
0&5 u|*}5
(
Qu&y + QS combination 6 *+. Other areas are
covers bigger area than the scope of 0&
6 *+ rule.
covered by Qû"o sūtras to 0&
120
Qˆ *+5
6
For 0&- *+, we should expect four examples, i.e. 1.Qu&y + pu&y (ƒ^1Œ5), 2.Qu&y + ƒu&y
(0¹9o$'),( 3.Qu&y + xu&y ($| Ž*}y5), and 4.Qu&y + {u&y (TuF"%5).
[LSK] ƒ^1Œ5 «
ƒ^ + pŒ
ƒ^ ‘( Œ 6 « ~ ‘$5 ^ku^y %395 Q*S *‹T"3"' (
6.1.87 w &5
6 letters,
1.1.50 v"4›1 t%T'5 « (Q is $ÁG5, p is T"iZ5, thus, among the 0&
‘ which has $ÁlT" as v"4 is the closest.)
ƒ^1Œ
0¹" + ƒo$
0¹ ,
( o$ 6 « ~ ‘$5 ^ku^y %395 Q*S *‹T"3"' (
6.1.87 w &5
1.1.50 v"4›1 t%T'5 «
0¹9o$
Small questions:
Find examples in the combinations below.
1. Q + p 5. Q + ƒ
2. Q + 6. Q + ù
3. w + p 7. w + ƒ
4. w + 8. w + ù
Explain how T^%$%& with 1.1.70 T^%ÆB"iŠ « works.
6.1.87 w &5
6 has two Qû"os. What are they and how are they Qû"o?
121
Qˆ *+5 3) a. ƒ%& %•( ^%5 “$| Ž*}y5”
6 letter Q. In
In the third and fourth examples, xu&y and {u&y are replaced by the 0&
such cases, the Q should be followed by % (and i,( respectively. This additional operation is
(
told by 1.1.51 ƒ%& %^%5 «. This sūtra requires pT ( !" for nasalized vowels, which is taught in
the next sūtra.
[ !" `k ']( 1.3.2 6 $ pT «(

ƒô12›1 —44"*
In the original texts of grammar, a nasalized vowel is pT. (
( Q44"*
ƒô12 1 7/1 QS 1/1 6 (
$5 1/1 pT 1/1
• ƒô12 1 7/1 – The original grammar teaching taught by three sages. Q*q$%&1 â'/.
• (
QS 1/1 – #e"‹"%5 QS, ( indicating all the vowels. This is !/.
• 6
Q44"* 6
$5 1/1 – Q44"* $ is a nasal sound which was defined by 1.1.8
6
6 4"*o$"uS49›44"*
'Ó $5 «; this is an adjective to QS. ( Together, a nasalized vowel.
• (
[LSK] ƒô12 1 7/1 Q44"*

6 ( pT-( !$5 1/1 Š"T III/1
$5 1/1 QS 1/1 ( «
In the original teaching, a nasalized vowel is termed pT. (
The following śloka tells what exactly ƒô12 means.

#e3"5 *2u `k "*& ã"o12" w0'"Æ‡" «
q"T^6 "l9 0&1 ^"l ƒô12"5 #$/*TyT"5 -
•6 (
1) #e3"5 – Suffixes taught in sūtras. E.g., ™* ,• *&*4•, 'T^
k "*& – '"‹1Â% `k "*&. E.g., i• & «(

2) *2u `
3) wo12"5 – Substitutes taught in sūtras. E.g., u • 6
4) w0'"5 – Augments taught in sūtras.
6 «, %' • 6 ¨/¸"3"' «(
5) q"T^6 "l5 – Those which are enumerated in q"T^6 "l5. E.g., Q • n*u
6) 0&^"l5 – Those which are enumerated in 0&^"l5. E.g., nuT • 6 «
122
Qˆ *+5
6
[LSK] #*T!"44"* À"5 1/3 ^"*&4/3"5 1/3 «
How do we know that certain vowels in Z"$%&2"¥ are nasalized? This u|*§ is
describing nasalized vowels used in the Z"$%&2"¥. (These nasalized vowels) taught by
6
^"*&*4 (^"*&4/3"5) are the ones whose status of nasal is taught in the tradition (#*T!"44"* À"5).
Summarizing, nasal vowels in Z"$%& are as per teaching.
[*»å&/] #*T!" 1/1 w44"*

6 ( 31s" 6/3 T1 1/3 T‡9´"5 1/3 ^"*&*4#n|*T#9´"5 1/3 31 1/3 u&"y5 1/3 T1 1/3 #*T!"-
À' 1/1
6
*us3-w44"* Àut5 1/3 pe‡5y «
[n •'/Z"¤"4']( ^"*&*44" #9´"5 u&"y5 ^"*&4/3"5 (4.3.101 T14 #9´' «( 4.2.114 u|}"•5«) « 07^%†%9ô1
6 241
6
(#*T!"3T1 Q43" p*T #*T!") w44"* 6
À"5 Q44"* $q'yut5 *t «
[LSK] iÁ `k v-Qu&Ë4 3/1 ‹ 0 ƒí"3y'"&5 1/1 %1Ý5 1/1 %i395 6/2 !" 1/1 -
The %• of “%•^%5”, which is %1Ý with Q• in i• &, ( the 6th sūtra of '"‹1Â% `k s, is a name for % (
and i.( This #e"‹"% %• is used in the next sūtra.
‹( Q 3 ( Q u ( Q % ( Q »( i ( Q• & (
#e"‹"% %• is a !" for % (and i (
(
» ( and & are ( 1.3.3 ‹i•' «.
pT by ( Q• is pT by
( 1.3.2 ƒô12›1 —44"*
6 ( #e"‹"% is made
$ pT «.
( Q•. Remember the
by 1.1.71 w*o5 Q•14 pT" ‹ «. Of #e"‹"% %•, % (is w*o letter, and Q• pT is
LSK u|*§5 for '"‹1Â% `k "*&. “‹$"%"*os 6 Q$"%5 ƒí"%&"‡5y « iÁ'j1 T 6 p„!$5 «“
The #e"‹"% %• is used in the next sūtra.
123
[^*%n"s" `k ']( 1.1.51 ƒ%& %•( ^%5 «

( the place of xu&y or {u&y should become %•^%, that which is followed by % (or i,(
Q& in
respectively.
( %•^%5 1/1 «
ƒ5 6/1 Q& 1/1
3 words in the sūtra, no Q4u|6 *§ is required.
• ƒ5 6/1 – #"*T^*o$ is x.; in v"431 90" sé/. “In the place of xu&y and {u&y”.
• (
Q& 1/1 – #e"‹"%5 with ^ku-y &$"%5, indicating Qu&y, pu&y, ƒu&y by 1.1.69 Q&*6 oT ( uy&Š
y
S"#e35«.
• (
%•^%5 1/1 – #e"‹"%5 %•, the first letter is % (of ‹3u%» (« and the end pT letter is Q• (by 1.3.2
6
ƒô12›1 —44"* $ pT«)( of i• &«.( %•: ^%5 3¼"T ( 5 %•^%5 (115B) from which ^% is %• (% (and i)( is %•^%5.
( «
[LSK] “x” p*T 0 *`2T5 6/1 !" 1/1 p*T 0 ƒ´' 1/1
It is said that x is a !" for 30 u&ys, 18 xs and 12 {s by 1.1.69 Q&*6 o„u&yŠ S"#e35«
(
and u"*Ty$ “x{u&y39*'y‡5 "uÁ3Ñ u"Î' «”.
[LSK] T (
"4 1 7/1 (x{u&yŠ 6/1 v"4 1 7/1) 35 1/1 Q& 1/1 ( ‘u 0 #uTyT 1 III/1 «
5 1/1 %•^%5 1/1 4 1/1
( to be there, it should be as %•^%5, that
In the place of xu&ys and {u&ys, when Q& is
which is followed by % (or i.(
$| Ž + x*}
( *}
$| Ž Q 6 « ~ ‘$5 ^ku^y %395 Q*S *‹T"3"' (
6.1.87 w &5
6 letters, Q is the
y 5. Among the 0&
1.1.50 v"4›1 t%T'5 « (Q is $ÁG5, x is 'kqÕ
closest.)
( (*}
$| Ž Q% 1.1.51 ƒ%& %•( ^%5 « (When x/{u&y becomes Q&, ( it should be %•^%5.)
1.1.50 v"4›1 t%T'5 « (among % (and i ( in %•, % (is the closest to 9.)
$| Ž*}y
124
Qˆ *+5
Tu + {$"%5
( $"%5
Tu Q 6 « ~ ‘$5 ^ku^y %395 Q*S *‹T"3"' (
6.1.87 w &5
6 letters, Q is the
1.1.50 v"4›1 t%T'5 « (Q is $ÁG5, { is o•5. Among the 0&
closest.)
(
Tu Qi ( $"%5 1.1.51 ƒ%& %•( ^%5 « (When x/{u&y becomes Q&, ( it should be %•^%5.)
1.1.50 v"4›1 t%T'5 « (among % (and i ( in %•, i ( is the closest to {.)
TuF"%5
Other examples:
%"—sy35, '‹*sy5 â*sy5, u tT5y6 , ƒ§'&y5, Qq'&y5
Small questions:
( used to make #e"‹"% Q& of
Which & is ( ƒ%& %•( ^%5?
How does x stand for 30 u&ys?

Explain how #e"‹"% %• is made, and how it is used.
6
How do we get w44"* À', ( the status of Q44"*
6 $ for the letters?
125
(
Qˆ *+5 3) b. ^ku`y "* }' “‹% p‹”
‹%1 + p‹
(
‹% (Q3 p‹ (
6.1.78 ‘S9›3u"3"u5 « ~ Q*S *‹T"3"' with ( o16 25 '"4"' «(
1.3.10 3‡" û' Q4
After applying 6.1.78 ‘S9›3u"3"u5« at the end of ô, the next sūtra will be applicable:
[*u*q `k ']( 8.3.19 i9^5 2"$ÞŠ « ~ Q^ku3y 95 ô"t395 Z95 Q*2 *‹T"3"' (
( at
3/u ( the end of ô is optionally elided when Q2 follows.
(
6 7
ô Qu&y ( (
3/u Q2 (
1
i9^
(
i9^5 1/1 2"$ÞŠ 6/1 « ~ Q^ku3y 95 6/2 ô"t395 6/2 Z95 6/2 Q*2 7/1 *‹T"3"' 7/1
2 words in `k ; 5 words as Q4u|6 *§
• i9^5 1/1 – This is wo12.
• 2"$ÞŠ 6/1 – The name of a grammarian respected by ^"*&*4, in ¬+sé/ to a supplied
word “'T1”, resulting in “in the opinion of 2"$Þ-'*6 4”. That means, it is not the opinion
of Pā ini himself. Thus, this indicates that this sūtra is optional. One form is according
to 2"$Þ, another from is according to Pā ini.
• Q^ku3y 95 6/2 – From 8.3.17, Q^kuŠ
y undergoes uS4-*u^*%&"' (modification of number) to
match with “Z95”. Qu&y5 ^ku5y 3"8" T… Q^ku> (115B), T395«, those which have Qu&y as ^ku;y
adjective to Z95. This Q is Q*uq/3'"45, thus it stands for 18 u&ys as per 1.1.69
Q&*6 o„u&yŠ S"#e35«.
• (
ô"t395 6/2 – From Q*q$"% `k ' 8.1.16 ôŠ « with Tot*u*q; in v"431 90" sé/.
To be exact, it should be Q^ku5y u$"%5 3$"%š, TotŠ ôŠ Q•Š Qi5 i9^5«. However, the
process is simplified in the u|*§.
• Z95 6/2 – u S ( Z… (ID), T395 « u$"% and 3$"%; in v"431 90" sé/.
( 3S
• Q*2 7/1 – #e‹"% Q2; ( all the vowels and soft consonants; in ^% â'/
• (
*‹T"3"' 7/1 – From 8.2.108; in *us3 â'/.
126
Qˆ *+5
[LSK] Qu&yûk 3y 95 6/2 ô"t395 6/2 3u395 6/2 i9^5 1/1 u" 0 Q*2 7/1 ^%1 7/1 «
(
i9^ is optionally the substitute in the place of 3 and (
u when they are at the end of ô
and preceded by Qu&y, and followed by Q2. (
[LSK] ‹% p‹, ‹%*3‹ «
‹%1 + p‹
(
‹% (Q3 p‹ 6.1.78 ‘S9›3u"3"u5 « ~ Q*S *‹T"3"' (
( o16 25 '"4"' «(
1.3.10 3‡" û' Q4
‹% (Q p‹ 8.3.19 i9^5 2"$ÞŠ « ~ Z95 Q^ku3y 95 ô"t395 Q*2 *‹T"3"' (
Now, can 6.1.87 w &5
6 « be applicable between Q and p?
The answer is no, because of the Q*q$"%- `k 8.2.1 ^ku`y "* }' «(
‹% p‹
^“1 (when 8.3.19 i9^5 2"$ÞŠ « was not applied), ‹%*3‹ «
127
The next Q*q$"% sūtra teaches a very important concept in Pā inisūtras.
[Q*q$"% `k ']( 8.2.1 ^ku`y "* }' «(

From 8.2.1 onward, these sūtras are as though non-existent from the view of
previous sūtras.
( «
^ku`y 0 Q* }' 1/1
• ( Q*q$%& â'/.
^ku`y 0 – ^ku y with ì ( ( âÅ‡Ë ì ( «) = ^ku*y ¼4 in
• (
Q* }' 1/1 (
– (^ku y `k ”H", from the view of previous sūtras) 4 * }' Q* }', ( that which
does not exist; subjective complement to 2"¥ `k ' (
The QU"j"3/ can be divided into two sections.

1. ^"o- â"j"3/
The first seven ( â4)( chapters (Qj"3) and the first quarter (^"o) of the eighth chapter
is called ^"o â"j"3/, the section which constitutes seven chapters along with one quarter.
^"o14 ‹ uTyT 1 p*T ^"o" « ‹-RºÔ/*‹ '" 5 (SB)«
6
^"o" Q … â"j"3/ S ^"o â"j"3/ « $'yq"%3-T°7s '" 5 (KT)«
2. *`^"o/
The last three (*`) quarters (^"o) are called *`^"o/, the section which constitutes three
quarters, starting from this sūtra 8.2.1 to the end of QU"j"3/. (`3"&" ^"o"4" '"‹"%5 *`^"o/ « *Ò0-6
6
T°7s '" 5 (DT)«)
128
Qˆ *+5
This sūtra teaches two things:

1. From the view of ^"o â"j"3/, *`^"o/ is non-existent.
( #*T 0 *`^"o/ 1/1 Q* }" 1/1«
[LSK] ^"o â"j"3/' 2/1
For example, the sūtra 8.3.19 i9^5 2"$ÞŠ «, which belongs to *`^"o/, does not exist
from the view of 6.1.87 w &5
6 «, which is ^"o â"j"3/. Thus 6.1.87 does not see the occasion
for application because from its view “3”( is still there between “Q” and p” as in the above
example (“‹% p‹”).
2. Also, within the *`^"o/, a latter sūtra is non-existent from the view of a preceding sūtra.
( Q*^ 0 ^ku'
[LSK] *`^"±"' 7/1 ( #*T 0 ^%' 1/1
y 2/1 ( 2"¥' 1/1
( Q* }' 1/1
( «
For example, the sūtra 8.2.23 390"tŠ i9^5 « does not exist from the view of 8.2.7
4i9^5 #"*T^*o$"tŠ «, which appears before.
( •6
n0uT +
n0uT ( 6.1.68 ‹i ( õ"I9 o/•"yT ( *TŠ^|
6 ´ ‹i ( « ~ i9^5
n0u 4 T( ( 7.1.70 ƒ*0oS" uy4"'v"4›1 q"T95 « ~ 4'6 (
n0u"4 T( ( 6.4.14 Q• tŠ S"q"T95 « ~ ƒ^q"3"5 o/•5y
n0u"4 ( 8.2.23 390"tŠ i9^5 «
After arriving at the form “n0u"4”,( 8.2.7 4i9^5 #"*T^*o$"tŠ « cannot be applied
(
because from the view of 8.2.7, the sūtra 8.2.23 does not exist. Thus, there is still T from the
(
view of 8.2.7; hence the condition to apply the elision of 4 does not arise.
Another example of 8.3.19 i9^5 2"$ÞŠ «.

[LSK] *uŽ p‹, *uŽ*u‹ «
*uŽ9 + p‹
( p‹
*uŽ Qu ( 6.1.78 ‘S9›3u"3"u5 « ~ Q*S *‹T"3"' (
( o16 25 '"4"' «(
1.3.10 3‡" û' Q4
( p‹
*uŽ Q 8.3.19 i9^5 2"$ÞŠ « ~ Z95 Q^ku3y 95 ô"t395 Q*2 *‹T"3"' (
*uŽ p‹ «
^“1 (when 8.3.19 i9^5 2"$ÞŠ « was not applied), *uŽ*u‹ «
129
(
Qˆ *+5 4) u|*}- *+5 “$| Ž•$•'”
First, the definition of the technical term u|*} is introduced in the next sūtra.
[ !" `k ']( 1.1.1 u|*}%"o•S «(

o/• y w, and o/• y ¾, ç - These three letters are termed u|*}.
( «
( ¾S 1/1
u|*}5 1/1 wT 1/1
• u|*}5 1/1 – This is !".
• (
wT 1/1 – This is !/.
This T^%-$%& (making it T^%, T$"%5 ^%5 3¼"T)( is not meant for $"i*43' by 1.1.70
T^%ÆB"iŠ«. Since w is not heard in '"‹1Â% `k "*&, w is not in Q&. ( Thus 1.1.69 Q&*6 o„u&yŠ
S"#e35« is not applicable to w. 1.1.70 T^%ÆB"iŠ « is applicable only to what is applicable
( not for $"i*43'. This T is
to 1.1.69 Q&*6 o„u&yŠ S"#e35«. Therefore, wT is ( just for clarity.
• (
¾S 1/1 – This is also !/. This is a #e"‹"% indicating ¾ and ç.
( T^%, which is after T. ( (T"T 5/1
This ¾S is ( ^%5 1/1 T^%5 1/1 = ¾S)( This T^%-$%& (making it
T^%) is meant for $"i*43' (limiting the '"`) by 1.1.70 T^%ÆB"%Š«.
[LSK] wT 1/1 ( S 0 u|*} !5 1/1 Š"T III/1

( ¾S 1/1 ( -
(
w and ¾S are termed u|*}.
u|*} indicates prosperity. By placing that !" at the very beginning, Pā ini expresses
'¹i in the shortest manner.
130
Qˆ *+5
The next *u*q `k gives the “u|*}- *+ rule”.
[*u*q `k ']( (
6.1.88 u|*}%1*S « ~ wT ‘$5 ^ku^y %395 *‹T"3"' (
When Qu&y is followed by ‘S, ( u|*} is the substitute for the two.
5 7
Qu&y ‘S (
1
u|*}5
( ‘$5 1/1 ^ku^y %395 6/2 *‹T"3"' 7/1

u|*}5 1/1 ‘*S 7/1 « ~ wT 5/1 (

• ( This is wo12.
u|*}5 1/1 – u|*}- !" is given to w, ¾, and ç by 1.1.1 u|*}%"o•S «.
• ‘*S 7/1 – #e"‹"%5 ‘S; ( in ^% â'/ by 1.1.66 T*¼*ü*T *4oËU 1 ^kuŠ
y «
This ‘S is
• (
wT 5/1 – From 6.1.87 w &5«.
6 6
#"*T^*o$ is Q; in ^ku^y Í'/ by 1.1.67 T¼"*oe§%Š«
• ‘$5 1/1 – From an Q*q$"%- `k 6.1.84 ‘$5 ^ku^y %395«. This is qualifying the u|*}-wo125.
• ^ku^y %395 6/2 – From an Q*q$"%- `k 6.1.84 ‘$5 ^ku^y %395«, in v"431 90" sé/; “in the place of ^ku y
and ^%”.
• (
*‹T"3"' 7/1 (
– From the Q*q$"% `k 6.1.72 *‹T"3"' «.
( ‘*S 7/1 ^%1 7/1 u|*}5 1/1 ‘$"o125 1/1 Š"T III/1
[LSK] wT 5/1 ( «
u|*} is the one substitute in the place of both ^ku y and ^% when Qu&y precedes and ‘S (
follows, in the topic of *‹T".
1/1
6
[LSK] 0&"û"o5 «
6 told by 6.1.87 w &5
u|*} is Qû"o, an exception, to 0& 6 «.
Definition of Qû"o5
314 4"#"â1 39 *u*q%"%8T1 TŠ"û"o5 «
314 (ƒ„0Ë&) 4 Q#"â1 (*us31) 35 (Qû"o5) *u*q5 w%8T1 5 (*u*q5) TŠ (ƒ„0yŠ) Qû"o5 «
131
A *u*q is started in the topic where ƒ„0y is not Q#"â (where ƒ„0y is definitely #"â),
that *u*q is Qû"o.
6 *+ is ƒ„0y. Qû"o is exception within the
ƒ„0y5 is a general rule. In this case, 0&
general rule. In this case, u|*} *+ is Qû"o.
[LSK] $| Ž•$•' «( 0¹…•5 « o1uÂ

• 3y' «( $| Ž…BÁG' -
(
$| Ž + ‘$•
$| Ž ¾( $• (
6.1.88 u|*}%1*S « ~ wT ‘$5 ^ku^y %395 *‹T"3"' (
1.1.50 v"4›1 t%T'5 « (Q is $ÁG5, ‘ is T"iZ5, thus, among the u|*} letters,
¾ which has $ÁlT" as v"4 is the closest.)
$| Ž•$• (oneness with KIJ a)
0¹" + ,•
0¹ ç
( • (
6.1.88 u|*}%1*S « ~ wT ‘$5 ^ku^y %395 *‹T"3"' (
1.1.50 v"4›1 t%T'5 « (Q is $ÁG5, , is , 5, thus, among the u|*} letters,
( v"4 is the closest.)
ç which has $Ál9é' as
0¹…• (stream of Gangā)
o1u + ¾Â3y
o1u ¾( Â3y (
6.1.88 u|*}%1*S « ~ wT ‘$5 ^ku^y %395 *‹T"3"' (
1.1.50 v"4›1 t%T'5 « (Q is $ÁG5, ¾ is $ÁlT"iZ5, thus, among the u|*} letters,
¾ which has $ÁlT" as v"4 is the closest.)
o1uÂ (
• 3y' (lordship of devas)
$| Ž + çBÁG
( BÁG 6.1.88 u|*}%1*S « ~ wT ‘$5
$| Ž ç ( ^ku^y %395 *‹T"3"' (
1.1.50 v"4›1 t%T'5 « (Q is $ÁG5, ç is , 5 , thus, among the u|*} letters,
( v"4 is the closest.)
ç which has $Ál9é' as
$| Ž…BÁG (desire for, in KIJ a) (çBÁG = ƒBÁl" + €"‡Ë ³È)(
132
Qˆ *+5
Qˆ *+5 4) a. ‘e1qek’ 6 “ƒ^ •*T”
The next *u*q `k gives a variation of the “u|*}- *+ rule”.
[*u*q `k ']( 6.1.89 ‘e1qek’ 6 « ~ wT ^k( u^y %395 ‘$5 u|*}5 ‘*S *‹T"3"' (
( 6 beginning with ‘S, ( 2) ‘q q"T
u|*} is the substitute after Qu&y, when 1) p& q"T ( 6
beginning with ‘S, ( or 3) ùl ( wo12 follows.

1) 7 2) 7 3)
5 5 5 7
Qu&y ( 6
‘S ( p& q"T Qu&y ( 6
‘S ( ‘q q"T Qu&y ùl ( wo12
u|*}5 u|*}5 u|*}5
( ‘$5 1/1 u|*}5 1/1 ^ku^y %395 6/2 ‘*S 7/1 *‹T"3"' 7/1
‘*T-‘q*T-ù’ 6 7/3 « ~ wT 5/1 (

• ‘*T-‘q*T-ù’ 6 7/3 – #"*T^*o$ is ‘e1qekl ( in ^% â'/ (1.1.66 T*¼*ü*T *4oËU 1 ^kuŠ
y )
‘*T5 S ‘q*T5 S ùl ( S ‘e1qekl5 (ID), T1s 6 «
(
‘*T = p& 0T… (
+ *ŸT^ (u"þ) ( .6
p*ßŸT^… q"T*6 4oË2 1 « ‘*T is a $| ot word just to indicate p&-q"T
(
‘q*T = ‘q•7 u|}… + *ŸT^ (u"þ) ( .6
p*ßŸT^… q"T*6 4oË2 1 « ‘q*T is a $| ot word just to indicate ‘q-q"T
ùl ( is an wo12 given by the sūtra 6.4.132 u"‹ ùl ( «
• (
wT 5/1 – #"*T^*o$ is Q, in 5th case (1.1.67 T¼"*oe§%Š«)
6
• ^ku^y %395 6/2 – from an Q*q$"%- `k 6.1.84 ‘$5 ^ku^y %395«, in v"431 90" sé/; “in the place of ^ku y
and ^%”
• ‘$5 1/1 – From an Q*q$"%- `k 6.1.84 ‘$5 ^ku^y %395«. This is qualifying the u|*} -wo125.
• u|*}5 1/1 – From the previous sūtra 6.1.88 u|*}%1*S «
• ‘*S 7/1 – From the previous sūtra 6.1.88 u|*}%1*S «
‘*S qualifies only ‘*T and ‘q*T told in 7th case. And this is Qi-œ‹&,
( mentioning of letters.
Thus To"*o*u*q by (u"þ) 3*¼*µ*qÆo"o"u¢œ‹&1 « is applied. As a result, they are read together
(
as “‘—"±95 7/2 ‘e1qe95 7/2 (when p& and ( 6 which begin with ‘S follow)”.
‘q-q"T (
This ‘S is
• (
*‹T"3"' 7/1 – the Q*q$"% `k 6.1.72 *‹T"3"' (
133
( ‘—"±95 7/2 ‘e1qe95 7/2 ù*l 7/1 u|*}5 1/1 ‘$"o125 1/1 Š"T III/1
[LSK] Qu&"yT 5/1 ( «
(
After Qu&y, when ‘S-beginning ( 6 or ‘S-beginning
p& q"T ( ( ,6 or ùl ( follows, u|*} is
‘q q"T
the one substitute in the place of the Qu&y and the following letter.
(
Now, we are seeing three examples: 1. Qu&y + ‘S-beginning ( ,6 2. Qu&y + ‘S-(
p& q"T
( ,6 and 3. Qu&y + ùl ( wo12.
beginning ‘q q"T
[LSK] ƒ^ •*T III/1 «
( ,6 which is an Qû"o for
The first example shows the combination of Qu&y + p& q"T
( which tells that after Qu&y obtains in ƒ^ 0y, when ‘, , beginning q"T 6
6.1.94 ‘*™ ^%–^' «,
follows, ^%–^ is the one substitute in the place of the two.
ƒ^ + ‘*T
ƒ^ ¾( *T 6.1.87 w &5
6 « is negated by 6.1.88 u|*}%1*S «, which is again negated by
( Finally ^%–^' is
6.1.94 ‘*™ ^%–^' «. ( negated by 6.1.89 ‘e1qek’ 6 «
ƒ^ •*T (he comes close, or he attains)

[LSK] ƒ^ •qT1 III/1 «
(
This is also an Qû"o for 6.1.94 ‘*™ ^%–^' «.
ƒ^ + ‘qT1
ƒ^ ¾( qT1 6.1.87 w &5
6 « is negated by 6.1.88 u|*}%1*S «, which is again negated by
( Finally ^%–^' is
6.1.94 ‘*™ ^%–^' «. ( negated by 6.1.89 ‘e1qek’ 6 «
ƒ^ •qT1 (being near, it grows)
[LSK] #é…‹5«
This is an Qû"o for 6.1.87 w &5«.
6
u‹ ( #"^&1 + *Áu 3.2.64 u‹š «
u"‹ ( 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
(
#é + u"‹ ( + 2 2/3
#é + ùl ( w‹ ( + Q ( (
6.4.132 u"‹5 ùl ( « ~ nŠ # "%&' (3& (
becoming p$)(
#é + ù‹ ( + Q ( (
y ^' Q*S
6.1.108 K "%&"í « ~ ^ku–
#é…‹ ( + Q ( By 6.1.87 w &5«,
6 6 is #"â5, but debarred by 6.1.89 ‘e1qek’ 6 «~ u|*}5
0&5
#é…‹5 8.2.66, 8.3.15
(A calf which carries logs set on its shoulder to keep it in one place)
134
Qˆ *+5
[LSK] ‘—"±95 7/2 *$' 0( ? «

( is called oi$| e or ô$| e to examine the necessity of a certain word.
“~ p*T *$'?”
(
Here the ‘*S as an adjective to ‘*T and ‘q*T is questioned. ‘—"±95, ‘S with w*o as To"*o*u*q, in
( 6 and ‘q q"T
dual number, is to specify that the word made of p& q"T ( 6 should begin with ‘S (
to be applicable for this u|*} change.

[LSK] ƒ^1T5«
(
1) Non-‘S-beginning ( 6
word with p& q"T
ƒ^ + pT5 ( ´-#e3 in 1/1, or p& +
pT5, p& + ( i» (/$Ty*%/III/2, does not start with ‘S. (
ƒ^1T5 6.1.87 w &5

6 «
Because of the ‘—"±95 word, this case is excluded from 6.1.89 ‘e1qek’ 6 «.
( *oqT «(
[LSK] '" nu"4 #1
(
2) Non-‘S-beginning ( 6
word with ‘q q"T
( *&S (
‘q +
‘*q
( $Ty*%/III/1
‘*q + ™/
‘*q + S™ ( + T ( 6 $Ty*% S™ ( «
3.1.48 *&*.L6 M85
p*q + Q + T ( 6
7.4.1 &… Sõ^q"3" •€5 «
p*q + *q + QT ( 6.1.11 S*™ « ~ Ò1 Q—"o15 *ÒT/3Š
( QT (
p*q + q + 6.4.51 &1%*4*» «
( QT (
p*o + q + 8.4.54 Q8" 1 SSy « ~ —2 (
p*oqT ( w» (-w0'5 for ™ ( does not come because of 6.4.74 4 '"™3
( 901«
# + p*oqT (
#1*oqT ( 6.1.87 w &5
6 «
6
Because of the ‘—"±95 word, this case is excluded from 6.1.89 ‘e1qek’ «.
( *oqT (May
'" nu"4 #1 ( the revered one not grow)
Note that nuT,• 6 respectful way to address “you”, takes third person.
135
Qˆ *+5 4) b. u"*Ty$s for ‘e1qek’ 6 “Q“…*‹&/”
Some u"*Ty$s are adding more instances to bring u|*} as Qû"o.

6
The next u"*Ty$ makes a particular word “Q“…*‹&/”. This is 0&-Qû"o.
(u"*Ty$')( Q“" *‹Õ"'^6 û"4' «( ~ ^ku^y %395 ‘$5 u|*}5 *‹T"3"' (

After Q“, when ù*‹4/ follows, u|*} is the one substitute, to make Q“…*‹&/.
( ù*‹Õ"' 7/1
Q“"T 5/1 ( ƒ^ û"4' 1/1
( « ~ ^ku^y %395 6/2 ‘$5 1/1 u|*}5 1/1 *‹T"3"' 7/1
(

• (
Q“"T 5/1 6
– #"*T^*o$ is Q“ in ^Í'/ as per 1.1.67 T¼"*oe§%Š «)
• (
ù*‹Õ"' 7/1 y )
– #"*T^*o$ is ù*‹4/ in ^% â'/ (1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
• (
ƒ^ û"4' 1/1 – Additional enumeration.
[LSK] Q“…*‹&/ 1/1 41 " 1/1 «

Q“ + ù*‹4/
( *‹4/
Q“ ç By 6.1.87 w &5«,
6 6 is #"â5. But it is debarred by (u"þ) Q“" *‹Õ"'^6 û"4'«.(
0&5
Q“…*‹&/ 8.4.3 ^ku^y o"T ( !"3"'05 « ~ %s"8" 49 &5 '"4ô1
Q“…*‹&/ (An army which consists of 21,870 chariots, 21,870 elephants, 64,610 horses,
109,350 foot soldiers)
(u"*Ty$')( #" ‹9ñ9N1s •³1s 6 « ~ ^ku^y %395 ‘$5 u|*}5 *‹T"3"' (

After #, when these 5 words follow, u|*} is the one substitute.
6
The first 3 words are of 0&-Qû"o, the last 2 words are of ^%–^-Qû"o.
( ƒ‹-ùñ-ù*ñ-‘s-‘³1s 6 7/3 « ~ ^ku^y %395 6/2 ‘$5 1/1 u|*}5 1/1 *‹T"3"' 7/1
#"T 5/1 (

• (
#"T 5/1 6
– #"*T^*o$ is # in ^Í'/ as per 1.1.67 T¼"*oe§%Š «)
• ƒ‹-ùñ-ù*ñ-‘s-‘³1s 6 7/1 – #"*T^*o$ is ù‹-ùñ-ù*ñ-‘s-‘³, pT%1T%Ò) '" , in ^% â'/
Since there are 5 types of words in the u"*Ty$, there are five examples.
136
Qˆ *+5
[LSK] #…‹5 «
# + ù‹ (ù‹ ( to guess + •È)(
( ‹
#ç By 6.1.87 w &5«,
6 6 is #"â5. But it is debarred by (u"þ) #" ‹9ñ9N1s •³1s 6 «.
0&5
#…‹ (The best logician)
[LSK] #…ñ5 «
# + ùñ (# + u‹ ( to grow + ´ 8)
( ñ
#ç By 6.1.87 w &5«,
0&5
#…ñ (The one who has grown, an arrogant person)
[LSK] #…*ñ «
# + ù*ñ (# + u‹ ( to grow + *´4)(
( *ñ
#ç By 6.1.87 w &5«,
0&5
#…*ñ (growth, arrogance)
[LSK] # •s5 «
( •È)(
# + ‘s (ps +
# ¾( s (
By 6.1.88 u|*}%1*S«, u|*}5 is #"â". But it is debarred by 6.1.94 ‘*™ ^%–^' «.
( debarred by (u"þ) #" ‹9ñ9N1s •³1s 6 «.
Again ^%–^' is
# •s (sending somebody)
[LSK] # •³5 «
( Á3T)(
# + ‘³ (ps +
# ¾( ³ (
By 6.1.88 u|*}%1*S«, u|*}5 is #"â". But it is debarred by 6.1.94 ‘*™ ^%–^' «.
( debarred by (u"þ) #" ‹9ñ9N1s •³1s 6 «.
Again ^%–^' is
# •³ (one who is sent)
8
u‹ ( + ´ 3.2.102 *4é" « ~ nkT 1
ƒ‹ ( + T ( ( is in 3—"*o0&.)
u*S€*^3—"o/4" *$*T « ~ K %&' (u‹
ƒñ ( + T 8.2.31 ‹9 ñ5 « ~ ð*i
ƒñ ( + q 8.2.40 ðsÆ‡9qÌ›q «
ƒñ ( + ñ 8.4.41 U64" U65 « ~ Æ95
ƒ +ñ 8.3.13 ñ9 ñ1 i9^5 «
ù +ñ y o/•Ì›&5 «
6.3.111 Oi9^1 ^kuŠ
137
(u"*Ty$')( xT1 S T|T/3" '" 1 « ~ ^ku^y %395 ‘$5 u|*}5 *‹T"3"' (

When the word xT follows after Qu&y in a '" in which the Qu&y-ending word is in
6
T|T/3", u|*} is the one substitute. This is 0&-Qû"o.
( ^ku^y %395 6/2 ‘$5 1/1 u|*}5 1/1 *‹T"3"' 7/1

xT1 7/1 S 0 T|T/3" '" 1 7/1« wT 5/1 (

• xT1 7/1 – #"*T^*o$ is xT, in ^% â'/
• S 0 – also, connecting to the topic of 6.1.89.
• T|T/3" '" 1 7/1 – #"*T^*o$ is T|T/3" '" , in ^% â'/
6 5«
[LSK] Ó"Ty
6 4 xT5 (= 0T5) p*T Ó"Ty
Ó1 6 5«
6 + xT (x to go + ´)
Ó
Ó (
6 w% (T By 6.1.87 w &5«,
6 6 is #"â5, but is debarred by (u"þ) xT1 S T|T/3" '" 1 «.
0&5
^*%n"s" `k s 1.1.50 v"4›1 t%T'5 « and 1.1.51 ƒ%& %•( ^%5 « are required.
6
Ó"Ty (One who has gone easily)
[LSK] T|T/3" p*T *$' ?( ^%'Ty5 «
6
oi$| e is done for the word “T|T/3"”. The counter example is in $'yq"%3T°7s '" .
^%'š Q … xT5 p*T ^%'Ty5«
^%' + xT
( (T
^%' Q% 6.1.87 w &5«
6
Since the ^ku^y o is not in T|T/3", this u"*Ty$ is not applicable.
^%'Ty (An ultimately freed person)
138
Qˆ *+5
(u"*Ty$')( #u„T%$¬iu 4"&yo2"4"'|& 1 « ~ ^ku^y %395 ‘$5 u|*}5 *‹T"3"' (

After these 6 words, when the word “x&” follows, u|*} is the one substitute.
6
This is 0&-Qû"o.
( x&1 7/1 « ^ku^y %395 6/2 ‘$5 1/1 u|*}5 1/1 *‹T"3"' 7/1
#-u„T%-$¬i-u 4-x&-o2"4"' 6/3 (

• (
#-u„T%-$¬i-u 4-x&-o2"4"' 6/3 (
– #"*T^*o$ is #-u„T%-$¬i-u 4-x&-o24 (ID), in sé/,
6
which is considered to be ^Í'/. (1.1.67 T¼"*oe§%Š «)
• y «)
x&1 7/1 – #"*T^*o$ is x& (debt), in ^% â'/ (1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
Now we are to see 6 examples. They are all '" s.
[LSK] #"&y', ( u„T%"&y', ( pe"*o«
# + x&
(
# w% (& By 6.1.87 w &5«,
6 6 is #"â5, but debarred by (u"þ) #u„T%$¬iu 4"&yo2"4"'|&«1 .
0&5
#"&y (Big debt)
u„T% + x&
u„T% (w% (& By 6.1.87 w &5«,
0&5
u„T%"&y (debt because of a grown calf)
$¬i + x&
$¬i ( w% (& By 6.1.87 w &5«,
0&5
$¬i"&y (debt for blanket)
u 4 + x&
(
u 4 w% (& By 6.1.87 w &5«,
0&5
u 4"&y (debt for clothes)
x& + x&
(
x& w% (& By 6.1.87 w &5«,
0&5
x&"&y (debt for debt)
o2 + x&
(
o2 w% (& By 6.1.87 w &5«,
0&5
o2"&y (o2 x&"*4 3*¼4, ( a name of a place where there are ten types of lakes, etc.)
139
Qˆ *+5 4) c. ƒ^ 0"y”*T q"T… “#"•y*T”
(
In the section of ‘$5 ^ku^y %395, now the topic of ^%–^' starts. The first *u*q `k we study
under this topic is 6.1.91 ƒ^ 0"y”*T q"T… «, which requires the knowledge of two !"s: ƒ^ 0y5
and q"T56 . Thus we shall see two !" `k s before the *u*q `k .
The next sūtra gives the ƒ^ 0y !".
[ !" `k ']( 1.4.58 ƒ^ 0"y5 *¨3"3901 « ~ #"o35

The 22 #"*os, when there is a meaningful connection to action, are termed ƒ^ 0"y5.
ƒ^ 0"y5 1/3 *¨3"3901 7/1 « #"o35 1/3

• ƒ^ 0"y5 1/3 – This is !".
• *¨3"3901 7/1 – *¨33" 3905*¨3"3905 (3T) connection with action; in *T â'/
• #"o35 1/3 – from 1.4.57 #"o35 « ; 22 particles starting with #. #-2²5 w*o5 31s" T1 #"o35
(116B) « They are enumerated in LSK u|*§. This is !/.
[LSK] #"o35 1/3 *¨3"3901 7/1 ƒ^ 0y !"5 1/3 Š56 III/3 «

1. # 2. ^%" 3. Q^ 4. '( 5. Q4 6 6. Qu
7. *4 9( 8. *4% ( 9. Ê ( 10. Ê% ( 11. *u 12. w™ (
13. *4 14. Q*q 15. Q*^ 16. Q*T 17. 6 18. ƒo (
19. Q*n 20. #*T 21. ^*% 22. ƒ^
[LSK] ‘T1 1/3 #"o35 1/3 -
Note that 18. ƒo ( is o$"%"t word. If “ƒT”( is written, that is because of 8.4.56 u"u "4«1 .
9
(
Meaning-wise, there is no distinction between *4 and (
*4%,( and Ê and Ê%.( That is why #"*o0& is thought
to be consisting of only 20 in some places. However, the reason *4% (and Ê% (should be counted separately
y "3T… «. This sūtra gives i"o12 for %1Ý in ƒ^ 0 y.
is to address the forms “*4i"3T1” and “Êi"3T1” by 8.2.19 ƒ^ 0Š
(
If there were only *4 and Ê , ( the (
can become %1Ý only by 8.2.66 6 75«, which is Q* } from the ”*U
—s9
y "3T… «. Thus *4% (and Ê% (are required.
of 8.2.19 ƒ^ 0Š
140
Qˆ *+5
The next sūtra gives q"T-6 !" to the q"Ts6 listed in q"T^6 "l.
[ !" `k ']( 1.3.1 nku"o39 q"Tu5 «

The group starting with nk, and like u" are termed q"T.6
nku"o35 1/3 q"Tu5 1/3 «

• nku"o35 1/3 – This is !/. nkš u"š nku… (ID) « w*oš w*oš wo/ (‘$21su|*§5) « nku… wo/ 31s" T1
nku"o35 (116B) « More will be explained below.
• q"Tu5 1/3 – This is !".
[LSK] *¨3"u"*S45 1/3 &"o35 1/3 q"T 6 !"5 1/3 Š56 III/3 -
Words indicating action, which start with nk, are termed q"Ts6 .
Even though the words may be nk, u", etc., they have to be *¨3"u"*S4, ( words
indicating action, only. From the words of sūtra, how do we arrive at “*¨3"u"*S45” as in the
u|*§? This is explained in *»å&/ .
nk has two meanings:
1. ( seen in q"T^6 "l, a *¨3"u"S/
nk §"3"' as
2. nk ^|*‡u/, a u"S/, a word indicating something which has *i¹ (gender) and uS4
(number)
u" also has two meanings:
1. u" 0*T0+4395 as seen in q"T^6 "l, a *¨3"u"S/
2. u" *u$®5, an QZ3 word
By defining q"T 6 with nk and u" together, what is common, *¨3"u"S/ is understood, and
what are uncommon, u"S/ and QZ3 are negated.
There are two q"T-6 !"-*uq"3$- `k s: One is this sūtra for the q"Ts6 in q"T^6 "l5, the other
sūtra is 3.1.32 4"±t" q"Tu5 « for derived q"Ts6 .
141
These two !"s: ƒ^ 0y5 and q"T56 are used in the next sūtra.
[*u*q `k ']( 6.1.91 ƒ^ 0"y”*T q"T… « ~ wT ^k( u^y %395 ‘$5 u|*}5 *‹T"3"' (
When Qu&y-ending ƒ^ 0y is followed by x$"%-beginning q"T,6 u|*} is the one substitute
for both Qu&y and x$"%.
ƒ^ 0"yT 5/1 ( ^ku^y %395 6/2 ‘$5 1/1 u|*}5 1/1 *‹T"3"' 7/1
( x*T 7/1 q"T… 7/1 « ~ wT 5/1 (

• (
wT 5/1 – from 6.1.87 w &5
6 «
#"*T^*o$ is Q; because this Q is Q*uq/3'"4 Q&, ( it is !" for Qu&y, all 18 types of Q,.
• (
ƒ^ 0"yT 5/1 6
– in the sense of 1.1.67 T¼"*oe§%Š «.
Since these two words are in '"4"*q$%&, Tot*u*q 1.1.72 314 *u*qÆotŠ « can be
(
applied; thus “Qu&"yt"T ƒ^ 0"yT”( or “Qu&y-Qt-ƒ^ 0"yT”( is understood.
• x*T 7/1 – x with T; ( the T^%$%& for 1.1.70 T^%ÆB"iŠ« to specify only short x;
indicating only x$"%s, short xs; in ^% â'/.
• q"T… 7/1 – in ^% â'/.
Since these two words are in '"4"*q$%& in 7th case, and one of them is Qi-œ‹&,
(
To"*o*u*q (u"þ) 3*¼*µ*qÆo"o"u¢œ‹&1« can be applied; thus “x$"%"o… q"T…” or “x$"%"*oq"T…”

is understood.
• u|*}5 1/1 – from 6.1.88 u|*}%1*S «. This is wo12
5 7
ƒ^ 0y Qu&y xT ( q"T 6
6
1
u|*}5
( ƒ^ 0"yT 5/1
[LSK] Qu&y-Qt"T 5/1 ( x$"%-wo… 7/1 q"T… 7/1 ^%1 7/1 u|*}5 1/1 ‘$"o125 1/1 Š"T III/1
( «
After Qu&y-ending ƒ^ 0y, when short x-beginning q"T 6 is following, u|*} is the one
substitute.
142
Qˆ *+5
[LSK] #"•y*T III/1 -

# + x•*T x• ( to go
6.1.87 w &5
6 « is #"â, but negated by:
(
# w% (•*T 6.1.91 ƒ^ 0"y”*T q"T… « ~ wT u|( *}5 ‘$5 ^ku^y %395 *‹T"3"' (
(
1.1.50 ƒ%& %^%5 «
#"•y*T
Side note 1:
Q: When q"T… can be understood by the presence of ƒ^ 0"yT, ( why is q"T… told?
A: To avoid optional #$| *Tn"u by xe$5.
Side note 2:
In ‘$"o12-Q*q$"%, ^ku y indicates the last letter of what is told in 5th case, and ^%
indicates the first letter of what is told in 7th case, because of the nature of *‹T", u&"y4"' (
Q*T2*3T5 *ü*q5, the close contact of letters.
Small questions:
What are the particles in #"*o0& which end with Qu&y?
What are the q"Ts6 in q"T$
6 92 which start with x$"%?
143
Qˆ *+5 5) ^%–^- *+5 “#1—T1”
We are still in the ‘$"o12 section. Now, ^%–^ topic starts.

The next sūtra specifies a narrow scope for ^%–^-wo12 as u|*}-Qû"o, which is further
6
negated by 6.1.89 ‘e1qek’ «.
[*u*q `k ']( 6.1.94 ‘*™ ^%–^' «( ~ wT ƒ^

( (
0"yT q"T… ^ku^y %395 ‘$5 *‹T"3"' (
When Qu&y-ending ƒ^ 0y is followed by ‘™-( beginning q"T,6 ^%–^ is the one substitute
for both Qu&y and ‘™.(
( « ~ wT 5/1
‘*™ 7/1 ^%–^' 1/1 ( ƒ^ 0"yT 5/1
( q"T… 7/1 ^ku^y %395 6/2 ‘$5 1/1 *‹T"3"' 7/1
(

• (
wT 5/1 – from 6.1.87 w &5
6 «
• (
ƒ^ 0"yT 5/1 6
– in the sense of 1.1.67 T¼"*oe§%Š «.
To those two words, Tot*u*q by 1.1.72 314 *u*qÆotŠ « can be applied; thus “Qu&"yt"T (
ƒ^ 0"yT”( or “Qu&y-Qt-ƒ^ 0"yT”( is understood.
• ‘*™ 7/1 – ‘™ ( with all u&ys, being Q*uq/3'"4; in ^% â'/.
• q"T… 7/1 – in ^% â'/.
Since these two words are in '"4"*q$%& in 7th case, and one of them is Qi-œ‹&,
(
To"*o*u*q (u"þ) 3*¼*µ*qÆo"o"u¢œ‹&1« can be applied; thus “‘™"o… q"T…” or “‘™"*oq"T…” is

understood.
• (
^%–'' 1/1 (
– This is wo12. This continues on as Q4u|6 *§ until 6.1.100. ^%Š –^ ^%–^' (6T)«
in the place of the ^ku y and ^%, only the one form, ^% will remain.
5 7
ƒ^ 0y Qu&y ‘™ ( q"T 6
6
1
^%–^' (
144
Qˆ *+5
( ƒ^ 0"yT 5/1
[LSK] wT 5/1 ( ‘™-( wo… 7/1 q"T… 7/1 ^%–^' 1/1
( ‘$"o125 1/1 Š"T III/1
( «
After Qu&y-ending ƒ^ 0y, when ‘™-( beginning q"T 6 is following, ^%–^ is the one
substitute.
Since ‘™ ( contains two letters, and what precedes is only Qu&y, we can see two
examples.
[LSK] #1—T1 «
# + ‘—T1 ‘—|7 • o/â… (1A) to shine
6.1.87 w &5
6.1.88 u|*}%1*S «. But further negated by:
( ‘ + —T1
#+ 6.1.94 ‘*™ ^%–^' «(
[LSK] ƒ^9s*T «
ƒ^+ ,s*T ƒs• o"‹1 (1P) to burn
6.1.87 w &5
6.1.88 u|*}%1*S «. But further negated by:
( , + s*T 6.1.94 ‘*™ ^%–^' «(
ƒ^ +
Small questions:
1. Explain the –^* *} of ƒ^ •*T and ƒ^ •qT1.
6 6.1.88 u|*}%1*S, 6.1.89 ‘e1qek’ ,6

2. Explain the ƒ„0y/Qû"o relationships among 6.1.87 w &5,
its u"*Ty$s, 6.1.91 ƒ^ 0"y”*T q"T…, and 6.1.94 ‘*™ ^%–^'. (
145
6
Qˆ *+5 5) ^%–^- *+5 a. 2$˜"*os 6 “2$+5”
The next rule of ^%–^ is taught by a u"*Ty$. To understand this, we first need to know
*» !", which is taught by the next !" `k .
[ !" `k ']( 1.1.64 QS9›•"*o *» «

A portion which starts with the last vowel is termed *».
‹i ( QS ( ‹i ( QS ( ‹i (
*»
QS5 6/1 Q•"*o 1/1 *» 1/1 «

• QS5 6/1 – #e"‹"%5 QS, ( “—"T… ‘$uS4' (singular
( to denote the class)” in *4q"y%&1 sé/;
“among the vowels”
• Q•"*o 1/1 – This is !/.
( 'j1 0) Q•5 1/1 w*o5 1/1 3Š 6/1 TT 1/1
(QS"' 6/3 ( Q•"*o 1/1 (116B)
That whose beginning is the last among the vowels.

• *» 1/1 – This is !".
[LSK] QS"' 6/3 ( *»- !' 1/1
( 'j1 0 35 1/1 Q•5 1/1 5 1/1 w*o5 1/1 3Š 6/1 TT 1/1 ( Š"T III/1
( «
For whom the beginning letter is the last vowel among the vowels, that is termed *».
For example, of “'4 ”,( *» portion is “Q ”.(
'( Q (
QS 4( Q (
QS (
*»
Small questions:
1. Tell *» of the following:
T, wT"', (
146
Qˆ *+5
Now, *» !" is used in the next u"*Ty$, which also teaches ^%–^'. (
(u"*Ty$')( 2$˜"*os 6 ^%–^ u"Î' «(

With reference to 2$+ 6 etc., for complying with those forms, ^%–^ should be told.
‹i ( QS ( ‹i ( QS ( ‹i (
*»
( u"Î' 1/1
2$+-6 w*os 6 7/3 ^%–^' 1/1 ( «
3 words in the `k , other words are to be understood by the context.

• 2$+-6 w*os 6 7/3 – 2$+56 w*o5 31s" T1 2$˜"o35 (116B), T1s 6 «, in *us3 â'/.
2$˜"*o10 is the name of a 0&, a group of words starting with 2$+.6
• (
^%–^' 1/1 – This is subject.
• u"Î' ( 1/1 – This is subjective compliment.
( (^%–^' 1/1
[LSK] TT 1/1 ( ) S 0 »1 5 5/1 (Q*S 7/1 ^%1 7/1)«
( ^%, ^%–^' also
In the place of *» as ^ku y and QS as ( happens.
[LSK] 2$+56 1/1 «
The first example: The *» part of 2$ is the last Q11. In the place of the *» and Q of Q+,6
(
^%–^ should be the substitute, which is justified by (u"þ) 2$˜"*Ts 6 ^%–^ u"Î' «.
2$ + Q+ 6
2$ ( + Q + + 6 By 6.1.101 Q$5 u&Ë o/qy5 «, o/qy5 is #"â5, but it is negated by
(u"þ) 2$˜"*Ts 6 ^%–^ u"Î' «(
2$+ 6 (the well of the place called 2$)
[LSK] $$y +56 1/1 «
10
(
2$˜"*o-0& contains: 2$+56 (2$"4"' Q+56 ), $$y +56 ($$y Š Q+56 ), /'t5 ( /Ä5 Qt5) $1 2u12 1 (in the
(
sense of the division of hair), '4/s" ('4 5 s"), ^Th*i5 (^T4 Qh*i5 3*¼4, ( to whom everyone salutes),
etc.
11
Since there is no letter after the last Q, the Q is not called w*o, after which something else should
(
follow. However, by the ^*%n"s" `k ' 1.1.20 ( Q is considered to be w*o. This is called
w±tuo1$*¼4 «,
Zô1*2u "u.
147
$$y + Q+ 6
$$y ( + Q + + 6 By 6.1.101 Q$5 u&Ë o/qy5 «, o/qy5 is #"â5, but it is negated by
(u"þ) 2$˜"*Ts 6 ^%–^ u"Î' «(
$$y + 6 (jujube berry)
[LSK ] '4/s" 1/1 «

( s"
'4 +
( + s"
'4 + (u"þ) 2$˜"*Ts 6 ^%–^ u"Î' «(
'4/s" (intellect)
( (= 2$˜"*o0&5 1/1) «
[LSK] w$| *T-0&5 1/1 Q3' 1/1
This 2$˜"*o0& is a group in which the number of its members is not fixed
(w$| *T0&).
There are two types of 0&s (groups).
1. Closed group (0& in general)
The general type of 0& is a group which has fixed number of members. For example,
the members of u"y*o0& are decided and fixed by Pā ini. No more members can be added.
2. Open-ended group (w$| *T0&)
A special type of 0& is called w$| *T0&. w$| *T0& is a group which does not have
fixed number of members. If there is a word in which the same operation is seen, and no
other sūtra or vārtika addresses the form, the word can join the 0&. €%"*o0& and S"*o0& are
examples for this type of 0&.
[LSK] '"TyÁ¸5 -
This is a word added to 2$˜"*o0&
'|T + QÁ¸
( Q + Á¸
'|T + By 6.1.101 Q$5 u&Ë o/qy5 «, o/qy5 is #"â5, but it is negated by
(u"þ) 2$˜"*Ts 6 ^%–^ u"Î' «(
'|TÁ¸
'|TÁ¸1 nu5 '"TyÁ¸5 ('|TÁ¸ + *™ + Q&)( 3
k 5y
148
Qˆ *+5
Qˆ *+5 5) ^%–^- *+5 b. ,'"™9š “*2u1*‹”
[*u*q `k ']( (
6.1.95 ,'"™9š « ~ wT ‘$5 ^ku^y %395 ^%–^' ( *‹T"3"' (
( w™ ( follow Qu&y, ^%–^' is
When ,' or ( the ‘$"o12.
5 7
Qu&y (
,'/w™ (
1
^%–^' (
(
,'-w™95 7/2
( ‘$5 1/1 ^ku^y %395 6/2 ^%–^' 1/1
S 0 « ~ wT 5/1 ( (
*‹T"3"' 7/1
• (
,'-w™95 ( w™ ( S ,'"™… (ID), T395 «; in ^% â'/.
7/2 – ,' S
• S 0 – This is to connect to the previous sūtra.
( ^%1 7/1 ^%–^' 1/1

[LSK] ,*' 7/1 w*™ 7/1 S 0 wT 5/1 ( ‘$"o125 1/1 Š"T III/1
( «
( w™ ( is following after Qu&y, ^%–^ is the one substitute.
When ,' or
( u|*}-Qû"o.
The combination of Qu&y + ,' is
( 0 4'5 0 «
[LSK] *2u"39' 4/1,
*2u"3 + ,' (
( ,+'(
*2u"3 + By 6.1.87 w &5«,
6 6 is #"â,
0&
6 is negated by 6.1.88 u|*}%1*S«. This is also negated by
0&
6.1.95 ,'"™9š« ~ ^%–^' (
*2u"39' (
149
The next example is about the combination of Qu&y + w™.(

To complete this form, we need the next sūtra.
[LSK] *2u 8/1 ‘*‹ II/1 -
(
(‹1) *2u + w™ ( + p*‹ (= p& 0T… + i9» (/$Ty*%/II/1)
This is the starting point.
Among these three words, there are two *+ applicable at the same time:
1. Between Q of *2u and w™,( u&yo/• y is possible.
6 is possible.
2. Between w™ ( and p*‹, 0&
Since $"3y of ƒ^ 0y and q"T 6 is considered to be Qt%¹, the $"3y of the more intimate
relationship12, 0&
6 is done first.
*2u + ‘*‹ 6 « ~ ‘$5 ^ku^y %395 Q*S *‹T"3"' (

6.1.97 w &5
6 ‘$"o12 ‘ is considered to be w™,( and 6.1.95
Now, with the help of the next sūtra, 0&
( applied.
,'"™9š « ~ ^%–^' is
*2u1*‹ 6.1.95 ,'"™9š « ~ wT ^k( u^y %395 ‘$5 ^%–^' ( *‹T"3"' (
12
q"Tk^ 0y395 $"3y't%¹' «(
150
Qˆ *+5
The next sūtra is an Q*To12 `k . Q*To12 is the extension of q'y from one entity to
another.
[Q*To12 `k ']( 6.1.85 Qt"*ouí « ~ ‘$5 ^ku^y %395

The ‘$"o12 is considered to be the end (Qt) of ^ku y and the beginning (w*o) of ^%.
w™ ( (
p& 0T…
(ƒ^ 0y5) (q"T56 )
‘
6
(‘$"o125 0&5)
This Q*To12 `k 6.1.85 Qt"*ouí « allows you to treat ‘$"o12 ‘ as w™ ( with ƒ^ 0y-q'y,
(
and p& with q"Tq6 'y.
By looking at the ‘ as w™ ( by this sūtra, 6.1.95 ,'"™9š « is applied.
Qt"*ouT 0( S 0 « ~ ^ku^y %395 6/2 ‘$5 1/1

• Qt-w*ouT 0( – Qt5 S w*o5 S Qto/ 1/2 (ID) «
( TÞ
Qt"*o8"' 3/2 ( Qt"*ouT 0( « 5.1.115 T14 TÞ
6 ' 1/1 6 *¨3" S1Ò*T5 « “like, same as Qt and w*o”.
• S 0 – This is to connect to the previous sūtra.
( ‘$"o125 1/1 5 1/1 ^kuŠ

[LSK] 35 1/1 Q3' 1/1 y 6/1 QtuT 0( ^%Š 6/1 w*ouT 0( «
In this section (from 6.1.84 to 6.1.111), the ‘$"o12 is like the end of ^ku y and the
beginning of ^%.
6
This sūtra comes right after 6.1.84 ‘$5 ^ku^y %395«, which means that all the ‘$"o12s (0&,
y ^, o/•,y etc.,) should be treated as Qt of ^ku y and w*o of ^%.
u|*}, ^%–^, ^ku–
151
[LSK] *2u1*‹ -
*2u + w™ ( + p*‹
*2u + ‘*‹ 6 « ~ ‘$5 ^ku^y %395 Q*S *‹T"3"' (
6.1.87 w &5
y QtuT. (
With the help of 6.1.85 Qt"*ouí «, ‘ is seen as w™,( ^kuŠ
( ‘ + *‹
*2u + 6.1.95 ,'"™9š « ~ ^%–^' (
*2u1*‹
Side note:
( .6 In that case, u|*} by 6.1.89 ‘e1qek’ 6 « is
If we take ^%Š w*ouT, ( ‘ is seen as p& q"T
possible. However, this does not happen. This is because 6.1.89 ‘e1qek’ 6 «, being
6
^%Æ"oû"o, can negate only 6.1.94 ‘*™ ^%–^', ( but not the next sūtra 6.1.95 ,'"™9š«, due to
6
the ^*%n"s" “^%Æ"oû"o" (
Q4t%"4 *uq/4 (
R"qt1 (
49§%"4 (^þ)”
6 then o/• y is u|*} by 6.1.88

In the case of Q± + w™ ( + ùñ", it becomes Q± + ,ñ" by 0&,
u|*}%1*S«, but it is negated by 6.1.95 ,'"™9š«. Thus Q±9ñ" is the final form. This is u|*}-Qû"o.
6 then o/• y is #"â by

In the case of Q± + w™ ( + x•*T, it becomes Q± + Q•y*T by 0&,
6.1.101 Q$5 u&Ë o/•5y «, but it is negated by 6.1.95 ,'"™9š«. Thus Q±•y*T is the final form.
This is o/•-y Qû"o.
152
Qˆ *+5
Qˆ *+5 6) o/• y- *+5 “o•e"*%5”
[*u*q `k ']( 6.1.101 Q$5 u&Ë o/•y5 « ~ Q*S ‘$5 ^ku^y %395 *‹T"3"' (
When Q$ ( is followed by u&y QS, ( o/• y is the ‘$"o12.
5 7
Q$ ( u&y QS (
6
1
o/• y
(
Q$5 5/1 u&Ë 7/1 o/•5y 1/1 « ~ Q*S ^ku^y %395 6/2 ‘$5 1/1 *‹T"3"' 7/1
• Q$5 5/1 – #e"‹"%5 Q$,( this is Q*uq/3'"4 Q$,( thus it brings all its u&ys; in ^ku^y Í'/.
• 6 "Š#3É u&y'«;( in ^% â'/.
u&Ë 7/1 – u&y is a !" defined as 1.1.7 TÞ
• o/•5y 1/1 – This is ‘$-wo12, one substitute, in the place of ^ku y and ^%.
(
[LSK] Q$5 5/1 u&Ë 7/1 Q*S 7/1 ^%1 7/1 ^ku^y %395 6/2 o/•5y 1/1 ‘$"o125 1/1 Š"T III/1 «
( following Q$,( o/• y is the ‘$"o12 in the place of ^ku y and ^%.
When u&y QS is
Because { is not seen very often, i•*6 }"t$…'o6 /$"% gives only four examples.
[LSK] o•e"*%5 «
( the demons) Q*%5 (enemy) p*T o•e"*%5 (6T) «
o•e"4"' (of
o•e + Q*%
( w + *%
o•e + By 6.1.87 w &5«,
6 6 is #"â, but negated by
0&
6.1.101 Q$5 u&Ë o/•5y «
1.1.50 v"4›1 t%T'5 «
o•e"*%
153
[LSK] ./25 «
*.3"5 (of LakJmī) 25 (lord) p*T ./25 (ViJ u) (6T)«
./ + 2
(
.+ +2 ( #"â, but negated by
By 6.1.77 p$9 3&*S«, 3& is
6.1.101 Q$5 u&Ë o/•5y «
1.1.50 v"4›1 t%T'5 «
./2
[LSK] *uŽko35 «
*uŽ95 (of ViJ u) ƒo35 (rise) p*T *uŽko35 (manifestation of ViJ u, or sunrise) (6T) «
*uŽ 6 + ƒo3
( ù + o3
*uŽ + ( #"â, but negated by
By 6.1.77 p$9 3&*S«, 3& is
6.1.101 Q$5 u&Ë o/•5y «
1.1.50 v"4›1 t%T'5 «
*uŽko3
[LSK] ‹9TQ$"%5 -
‹9T56 (of priest) x$"%5 (x sound) p*T ‹9TQ$"%5 (6T) «
‹9T| + x$"%
( 9 + $"%
‹9T + ( #"â, but negated by
By 6.1.77 p$9 3&*S«, 3& is
6.1.101 Q$5 u&Ë o/•5y «
1.1.50 v"4›1 t%T'5 «
‹9TQ$"%
Small questions:
What is the difference among ‘*™ ^%–^', ( 2$˜"*o0&, and ,'"™9š?
154
Qˆ *+5
y ^- *+5 “‹%1›u”
Qˆ *+5 7) ^ku–
[*u*q `k ']( 6.1.109 ‘™5 ô"t"o*T « ~ ^ku5y ‘$5 ^ku^y %395 *‹T"3"' (
When ô"t ‘™ ( is followed by a short Q, ^ku–
y ^ is the ‘$"o12.
5 7
ô ‘™ ( QT (
6
1
y ^
^ku–
( Q*T 7/1 « ~ ^ku5y 1/1 ^ku^y %395 6/2 ‘$5 1/1 *‹T"3"' 7/1
‘™5 5/1 ô"t"T 5/1 (

• ‘™5 5/1 – #e"‹"%5 ‘™;( in ^ku^y Í'/.
• ô"t"T (5/1 – ôŠ Qt5 ô"t5 (6T), T¼"T «;
( in ^ku^y Í'/.
• Q*T 7/1 – #"*T^*o$ is QT, ( Q with T; ( the T^%$%& for 1.1.70 T^%ÆB"iŠ« to specify only
short Q; in ^% â'/.
• y ^'. (
^ku5y 1/1 – From 6.1.107 Q*' ^ku5y «. This is understood as ^ku–
This is ‘$-wo12, one substitute, in the place of ^ku y and ^%.
( ‘™5 5/1 Q*T 7/1 ^%1 7/1 ^ku–

[LSK] ô"t"T 5/1 ( ‘$"o125 1/1 Š"T III/1
y ^' 1/1 ( «
After ‘™ ( at the end of ô, when short Q follows, ^ku–
y ^ is the substitute for both ^ku y
and ^%.
[LSK] ‹%1›u -
( protect + i9» (/$Ty*%/II/1)
(‹1) ‹%1 + Qu (Qu to
‹% (+ ‘ + u ( #"â, but negated by
By 6.1.78 ‘S9›3u"3"u5«, Q3 is
6.1.109 ‘™5 ô"t"o*T « ~ ^ku5y
‹%1u
155
[LSK] *uŽ9›u «
*uŽ9 + Qu
( ,+u
*uŽ + ( #"â, but negated by
By 6.1.78 ‘S9›3u"3"u5«, Qu is
6.1.109 ‘™5 ô"t"o*T « ~ ^ku5y
*uŽ9u
Note that Quœ‹ (›) is used to indicate the existence of Q, which is now gone. It is not
compulsory to use Quœ‹ at all. It is used totally according to the writers’ preference. Quœ‹
should not affect pronunciation in any manner. In fact, Quœ‹ was introduced only recently.
Some people call this sandhi “Quœ‹- *+” but this is not recommended for a grammar
student.
156
Qˆ *+5
Qˆ *+5 8) #$| *Tn"u- *+5 “09Qœ'”(
From here onward until the end of the QS-( *+5 section, #$| *Tn"u- *+ is taught.
“#$| *Tn"u” means “being as it is, without modification”, and is prohibition of *+
change where QS-( *+ can happen.
[*u*q `k ']( 6.1.122 uy` *un"s" 095 « ~ #$| e" ‘™5 ô"tŠ Q*T *‹T"3"' (
In both i9$ and u1o, when ô"t ‘™-( ending 09-2² is followed by a short Q, there is
optional #$| *Tn"u (being as it is without *+ change).
6 7
ô ‘™ ( of 09 QT (
1
Optional
#$| *Tn"u (no *+)
(
uy` 0 *un"s" 0 095 6/1 « ~ #$| e" 3/1 ‘™5 6/1 ô"tŠ 6/1 Q*T 7/1 *‹T"3"' 7/1
• uy` 0 – uy + ì ( by 5.3.10 âÅ‡Ë ì«;
( everywhere, in Vedic as well as non-Vedic
literature.
• *un"s" 0 – optional; if #$| *Tn"u is not taken, another ^“ is by 6.1.109 ‘™5 ô"t"o*T «.
• 095 6/1 – 09-2²; in v"431 90" sé/.
• (
#$| e" 3/1 – #"*T^*o$ #$| *T, unmodified form; in T|T/3" by (u"þ) #$| e"*o8 ƒ^ û"4' «;
connected to nu*T, it means “stays as its own form”.
• ‘™5 6/1 – From 6.1.109 ‘™5 ô"t"o*T «; *un*´*u^*%&"' from 5th case to 6th case is done.
#e"‹"%5 ‘™;( in v"431 90" sé/.
By Tot*u*q, “‘™tŠ 095” is understood.
• ô"tŠ 6/1 or ô"t1 7/1 – From 6.1.109 ‘™5 ô"t"o*T « by 'Á¸k$‚ 6T0*T5, Q4u|6 *§ like the
jump of a frog; *un*´*u^*%&"' to either 6th case, as adjective to 095, or Q*q$%&1 7th case to
make “at the end of ô”.
157
• Q*T 7/1 – From 6.1.109 ‘™5 ô"t"o*T «; #"*T^*o$ is QT, ( the T^%$%& for 1.1.70 T^%ÆB"iŠ«
to specify only short Q; in ^% â'/.
[LSK] i9$1 7/1 u1o 1 7/1 S 0 ‘™"tŠ 6/1 095 6/1 Q*T 7/1 u" 0 #$| *Tn"u5 1/1 ô"t1 7/1 «
In non-Vedic and Vedic literature, the word 09 which ends with ‘™,( when it is
followed by short Q, optionally stays as its own form, at the end of ô.
[LSK] 09Qœ', ( 09›œ' «(

( 09Qœ' (
095 Qœ' p*T
( Qœ + • 6
09 + ™ + 6 T°7s5
2.2.8 sé/ « ~ ^" 6 '" 5
09 + Qœ 6 q"T#6 "*T^*o$395« ~ $ (
2.4.71 ^9
Even after $ ( elision of ^6 , ( ^6 is
( seen 1.4.14 *â™t
6 ô' «( because of
( Thus 09 is considered to be ô.
1.1.62 #e3i9^1 #e3i“&' «.
09Qœ ( #"â, but negated by 6.1.109 ‘™5 ô"t"o*T «,
By 6.1.78 ‘S9›3u"3"u5«, Qu is
and further negated by 6.1.122 uy` *un"s" 095 « ~ #$| e" ô"t1 Q*T
^“1
09›œ 6.1.109 ‘™5 ô"t"o*T « ~ ^ku5y
Now, ô$| e for the word ‘™5, which was taken as Q4u|6 *§.
[LSK] ‘™-( QtŠ *$'?( *S`Ïuœ' «(
*S`"5 1/3 0"u5 1/3 3Š 5 *S`056 1/1 (116B) « One who has varied-coloured cows
( 09 + — ( 2.2.24 Q4$
*S` + — + 1 'Õô"‡Ë « ~ '" 5 RºÔ/*‹5
6
#"*T^*o$ !" by 1.2.46, R$ ( by 2.4.71
*S` + 0 6 1.2.48 09*¥397^ —y4Š « ~ •€5

( *S`Ïu' 1/1
*S`095 6/1 Qœ' 1/1 ( (6T) «
*S`0 6 + Qœ
( Qœ
*S`Ïu + 6.1.77 p$9 3&*S « ~ *‹T"3"' (
There is 09-2² at the end of the ô, but it is not ‘™.( Thus 6.1.122 uy` *un"s" 095 « does
not apply.
158
Qˆ *+5
Another ô$| e for the word ô"t1, which was taken as Q4u|6 *§.
[LSK] ô"t1 *$'?( 095 -
09 + ™* • or 09 + ™ (
09 + Q (
( ,+
0+ ( 6.1.110 ™* ™ 95 6/2 S 0 « ~ ‘™5 5/1 Q*T 7/1 ^ku^y %395 6/2 ‘$5 1/1 ^ku5y 1/1
(
After ‘™,( when short Q of ™* • and ™ follows, y ^ is the ‘$"o12.
^ku–
Here, , of 09 is not at the end of ô. ô"t is the ( the ^6 . ( Thus, 6.1.122 uy` *un"s"
of
095 « does not apply.
Of the combination “09 + Qœ”, there are three forms:

( 6.1.122 uy` *un"s" 095 « ~ #$| e"
1. 09Qœ' by
( 6.1.109 ‘™5 ô"t"o*T « ~ ^ku5y
2. 09›œ' by
( 6.1.123 Qu™ ( ê9»"34Š « ~ 095
3. 0uœ' by
To understand the sūtra 6.1.123 Qu™ ( ê9»"34Š «, we need to study two more
^*%n"s" `k s, which are coming next.
159
The next two ^*%n"s" `k s are from the section which teaches where the replacement
should happen.
[^*%n"s" `k ']( 1.1.55 Q41$"i ( *2T ( uyŠ«

1 "i ( or *2T. (
The substitute should be in the place of the all letters when wo12 is Q4$
1 "i 1/1
Q4$ (
( *2T 1/1 uyŠ 6/1 «
• 1 "i ( 1/1 – 4 ‘$5 Q4$
Q4$ 1 5 Qi ( 3Š 5 Q4$
1 5 (NT) « Q4$ 1 "i ( (116B) «
1 "i ( here.
This is an adjective to wo125. The wo12 which has many letters is called Q4$
• (
*2T 1/1 (
– 2$"%5 pT 3Š (
5 *2T (116B) «
( pT is
This is an adjective to wo125. The wo12 which has 2 as ( called *2T here.
(
• uyŠ 6/1 – in v"431 90" sé/ (1.1.49 sé/ v"431 90" «)
1 "i 1/1
[LSK *»å&/] Q4$ ( S 0 wo125 1/1 uyŠ 6/1 v"4 1 7/1 nuT5 III/2 «
( *2T 1/1
1 "i ( or *2T, ( should occur in the place of all the letters of

The substitute, which is Q4$
what is presented in 6th case.
This is Qû"o to 1.1.52 Qi9›•Š «.
1 "i ( –
Example 1: In the case of Q4$
%"' + *n (
%"' + ¾ ( ( «
7.1.9 QT5 5/1 *n 5 6/1 ¾ 1/1
( Q4$
¾ is ( replaced with the help of
1 "i.( Thus the whole *n is
1 "i ( *2T ( uyŠ «
1.1.55 Q4$
Example 2: In the case of *2T –(
( T* i
po' + • (
p+T ( ( «
5.3.3 po'5 6/1 p2 1/1
( *2T by
p2 is ( 1.3.3 ‹i•'«.( Thus the whole po' is
( replaced with the help of 1.1.55
1 "i ( *2T ( uyŠ «

Q4$
1 "i ( *2T ( uyŠ «.
The next ^*%n"s" `k is Qû"o to 1.1.55 Q4$
[LSK] p*T 0 #"â1 7/1« When this sūtra is applicable… (to be continued to the next sūtra.)
160
Qˆ *+5
[^*%n"s" `k ']( 1.1.53 *™í « ~ Qi5 Q•Š

The substitute should be in the place of the last letter when wo12 is *™T. (
*™T ( 1/1 S 1/1 « ~ Qi5 6/1 Q•Š 6/1

• (
*™T 1/1 (
– ™$"%5 pT 3Š (
5 *™T (116B) «
( called *™T here.
This is an adjective to wo125. The wo12 which has ™ ( as pT is (
• S 0 – This brings the context from the previous sūtra: 1.1.52 Qi9›•Š «
• Qi5 6/1 – #e"‹"%5 Qi;( any letter; in v"431 90" sé/ (1.1.49 sé/ v"431 90" «)
• Q•Š 6/1 – in v"431 90" sé/ (1.1.49 sé/ v"431 90" «)
( Q4$
[LSK] *™T 1/1 1 "i 1/1 (
( Q*^ 0 Q•Š 6/1 ‘u 0 Š"T III/1 «
Even though the wo12 has more than one letter, if wo12 is *™T, ( it replaces only the
last letter of what is presented in 6th case.
Example:
09 + Qœ
(
0 Qu™ ( + Qœ 6.1.123 Qu™ ( ê9»"34Š « ~ 095 ô"tŠ Q*S (the next sūtra)
1 "i,( only Q• Qi ( is replaced with the help of

Though Qu™ ( is Q4$
1.1.53 *™í « ~ Qi5 Q•Š
161
Summary of the section starting from 1.1.52 Qi9›•Š
1.1.52 Qi5 6/1 Q•Š 6/1 «

( S « ~ Qi5 6/1 Q•Š 6/1
1.1.53 *™T 1/1 Qû"o5
Qû"o5
1 "i 1/1
1.1.55 Q4$ (
( *2T 1/1 uyŠ 6/1 «
The word in 6th case indicates the letter to be replaced. The word in 1st case
indicates wo12.
As a general rule (ƒ„0y), the last letter is to be replaced by 1.1.52 Qi9›•Š«.

1 "i ( *2T ( uyŠ« is Qû"o to 1.1.52 Qi9›•Š«. By hearing Q4$
1.1.55 Q4$ 1 "i,( it is
( ‘$"i.( It is also understood the general rule is about ‘$"i.(
understood that *2T is
1 "i ( *2T ( uyŠ«. It is understood that *™T is
1.1.53 *™í « is Qû"o to 1.1.55 Q4$ ( Q4 1$"i.(
Otherwise, the 1.1.53 *™í « sūtra becomes useless because the case where ‘$"i ( wo12 is
replacing the last letter is already covered by 1.1.52 Qi9›•Š«.
162
Qˆ *+5
The next sūtra is also Qû"o to 6.1.109 ‘™5 ô"t"o*T «.
[*u*q `k ']( 6.1.123 Qu™ ( ê9»"34Š « ~ 095 ‘™5 ô"tŠ Q*S *‹T"3"' (
In the opinion of ê9»"34 '*6 4, Qu™ ( is the substitute for ô"t ‘™-( ending 09-2² when
followed by vowel.
6 7
ô ‘™ ( of 09 QS (
1
Optional
Qu™ (
(
Qu™ ( 1/1 ê9»"34Š 6/1 « ~ 095 6/1 ‘™5 6/1 ô"tŠ 6/1 Q*S 7/1 *‹T"3"' 7/1
• ( 2; thus it replaces only the last letter by 1.1.53 *™í «
Qu™ ( 1/1 – This is *™T wo1
• ê9»"34Š 6/1 – ê9»"34 is the name of a x*s grammarian. “'T1 7/1 (in the opinion)” can
be supplied; in ¬+sé/; “in the opinion of ê9»"34 '*6 4”. Option is implied.
• ô"tŠ 6/1 or ô"t1 7/1 – From 6.1.109 ‘™5 ô"t"o*T « by 'Á¸k$‚ 6T0*T5; *un*´*u^*%&"' to
either 6th case, as adjective to 095, or Q*q$%&1 7th case to make “at the end of ô”.
• 095 6/1 – 09-2²; in v"431 90" sé/.
#e"‹"%5 ‘™;( in v"431 90" sé/.
By Tot*u*q, “‘™tŠ 095” is understood.
• Q*S 7/1 – From 6.1.77 p$9 3&*S «; in ^% â'/.
[LSK] ô"t1 7/1 ‘™"tŠ 6/1 095 6/1 Qu™ ( 1/1 u" 0 Q*S 7/1 «
(
When QS follows, Qu™ ( is the substitute optionally in the place of the last letter of
09-2² which ends with ‘™ ( and at the end of ô.
163
[LSK] 0u"œ', ( 09›œ' «(

09 + Qœ
( Qu™ ( + Qœ
0+ 6.1.123 Qu™ ( ê9»"34Š « ~ 095 ‘™5 ô"tŠ Q*S
In the process of deciding which letter is to be replaced, first 1.1.52 Qi9›•Š « tells
1 "i ( *2T ( uyŠ « tells “all the letters” presented in 6th case,
“the last letter”. Then 1.1.55 Q4$
1 "i.( Finally, 1.1.53 *™í « tells “the last letter” because it is *™T. (
because the wo12 is Q4$
( Qu + Qœ
0+ 6
Q4R+i9^ (
of pT (1.3.3 ‹i•' «( and 1.3.9 TŠ i9^5«)
0u"œ 6.1.101 Q$5 u&Ë o/•5y «
Qu™-( Qn"u-^“1 (when Qu™ ( is not applied)

09›œ 6.1.109 ‘™5 ô"t"o*T «
Thus, for the combination of 09 + Qœ, there are three forms:

( 6.1.122 uy` *un"s" 095 « ~ #$| e" Q*T
1. 09Qœ' by
( 6.1.109 ‘™5 ô"t"o*T « ~ ^ku–
2. 09›œ' by y ^' (
( 6.1.123 Qu™ ( ê9»"34Š « ~ Q*S
3. 0u"œ' by
[LSK] ô"t1 *$'?( 0*u -

09 + *™ ( « ~ õ"T"*T^*o$"T (
4.1.2 €…— '…S …
#"*T^*o$ 09 + â'/ ‘$uS4' (
09 + p 6
Q4R+i9^ (
of pT (1.3.8 i2òT*}T1 « and 1.3.9 TŠ i9^5«)
( Qu +
0+ ( p 6.1.78 ‘S9›3u"3"u5 « ~ Q*S
09 is not at the end of ô.
164
Qˆ *+5
The next sūtra is just for achieving the form “0u1Œ” by negating the option of 6.1.123
Qu™ ( ê9»"34Š « ~ Q*S.
[*u*q `k ']( 6.1.124 pŒ1 S « ~ Qu™ ( 095 ‘™5 ô"tŠ *‹T"3"' (

09 + pŒ always becomes 0u1Œ.
6 7
ô ‘™ ( of 09 “pŒ”
1
Qu™ (
(
pŒ1 7/1 S 0 « ~ Qu™ ( 1/1 095 6/1 ‘™5 6/1 ô"tŠ 6/1 *‹T"3"' 7/1
• pŒ1 7/1 – A word “pŒ”; in ^% â'/.
• S 0 – Connects with the previous sūtra, 6.1.123 Qu™ ( ê9»"34Š «.
• ( 2; thus it replaces only the last
Qu™ ( 1/1 – From 6.1.123 Qu™ ( ê9»"34Š «; this is *™T wo1
letter by 1.1.53 *™í «
• ô"tŠ 6/1 or ô"t1 7/1 – From 6.1.109 ‘™5 ô"t"o*T« by 'Á¸k$‚ 6T0*T5; *un*´*u^*%&"' to
either 6th case, as adjective to 095, or Q*q$%&1 7th case to make “at the end of ô”.
• 095 6/1 – 09-2²; in v"431 90" sé/.
#e"‹"%5 ‘™;( in v"431 90" sé/. By Tot*u*q, “‘™tŠ 095” is understood.
(
[LSK] 095 6/1 Qu™ ( 1/1 Š"T III/1 pŒ1 7/1 «
When “pŒ” follows, Qu™ ( is the substitute in the place of the last letter of 09-2²
which ends with ‘™ ( at the end of ô.
165
( In this sūtra, there are two

| " Q*S *4e' «.
The next #$| *Tn"u sūtra is 6.1.125 ‚ 6T-#0ã
!"s: ‚ 6T and #0|ã. Thus we will study some sūtras relating to these !"s.
‚ 6T !" is defined by 1.2.27 ù$"i9› €o/•‚y 6T5«. Then, ‚ 6T is enjoined to certain
vowels by *u*q `k s in the section starting from 8.2.82 u"ÀŠ »1 5 ‚ 6T ƒo"§5 «. There, which
vowel should become ‚ 6T in what occasion is taught.
(
The next *u*q `k ' sūtra is one of such ‚ 6T-*uq"3$- `k s.
[*u*q `k ']( 8.2.84 %"}kT 1 S « ~ u"ÀŠ »1 5 ‚ 6T5 u"

In the act of calling from a distance, the *» of a sentence optionally becomes ‚ 6T.
( UT1 7/1 S 0 «~ u"ÀŠ 6/1 »1 5 6/1 ‚ 6T5 1/1 u" 0

%"T 5/1
• (
%"T 5/1 – #"*T^*o$ % (distance); in Q^"o4 1 ^Í'/; “from a distance”.
• (
UT1 7/1 – #"*T^*o$ UT (act of calling); ;1È Ðq"y (
3"' (1U) to call + ´ (n"u1); ¬9q4 1 is
understood; in *us3 â'/.
• S 0 – Connects with the Q*q$"% sūtra in this topic, 8.2.82 u"ÀŠ »1 5 ‚ 6T ƒo"§5 «.
• u"ÀŠ 6/1 – From 8.2.82 u"ÀŠ »1 5 ‚ 6T ƒo"§5«; in ¬+sé/ connected to »1 5.
• »1 5 6/1 – #"*T^*o$ *», defined by 1.1.64 QS9›•"*o *» «; in v"431 90" sé/.
• ‚ 6T5 1/1 – Defined by 1.2.27 ù$"i9› €o/•‚y 6T5«; this is wo12.
• u" 0 – Though this word is not in `k ^"l, '‹"n"³$"% explains that this is *u$® (optional).
(
[LSK] %"T 5/1 ¬9q4 1 7/1 u"ÀŠ 6/1 »1 5 6/1 ‚ 6T5 1/1 u" 0 -
( the wo12 in the place of *» of the sentence
In the act of calling from a distance, ‚ 6T is
optionally.
To understand the application of this sūtra in an example, the next ^*%n"s" `k is

required.
166
Qˆ *+5
( more than one letter, we need the next

When wo12 is •€, o/• y or ‚ 6T, and v"*44 is
^*%n"s" `k .
This ^*%n"s" `k is not in i•*6 }"t$…'o6 /.
[^*%n"s" `k ']( 1.2.28 QSš « ~ •€o/•y‚ 6T5

The word “QS5 6/1” should be added in u|*§ when wo12 is •€, o/• y or ‚ 6T, and the
place of replacement is not clear.
QS5 6/1 S 0 « ~ •€-o/•-y ‚ 6T5 1/1

• QS5 6/1 – #"*T^*o$ QS; ( in v"431 90" sé/.
• S 0 – connects with the previous sūtra.
• •€o/•‚y 6T5 1/1 – From the previous sūära 1.2.27 ù$"i9› €o/•‚y 6T5«.
( *uq/3T1 III/1 T` 0 “QS5 6/1” p*T 0 s t' 1/1

[SK] •€-o/•-y ‚ 6T-2²•5 3/3 3` 0 QS 1/1 ( ô' 1/1
( ƒ^*TéT1 III/1 «
Wherever a vowel is enjoined by the word “•€, o/• y or ‚ 6T”, in that place, 6th case-
ending word “QS5 6/1” is supplied.
For example, in the sūtra 8.2.84 %"o ( qkT 1 S « ~ ‚ 6T5 u"ÀŠ »1 5, because ‚ 6T5 is enjoined,
“QS5 6/1” is supplied and the meaning will become “‚ 6T is the substitute in the place of QS (
of *» of u"À when addressing from a distance.”
( «
‘*‹ II/1 n0u4 8/1
‘*‹ n0u"æ4 ( 8.2.84 %"o ( qkT 1 S « ~ ‚ 6T5 u"ÀŠ »1 5 QS5 6/1 u"
To decide which letter of *» is replaced by ‚ 6T, the ^*%n"s" `k 1.2.28 QSš « is applied.13
To decide which ‚ 6T letter should be the substitute, another ^*%n"s" `k 1.1.50
( by v"4 is $ÁG letter wæ.
v"4›1 t%T'5« is used. The closest to the v"*44 Q
When the optional ‚ 6T is not applied, no other change is required.
13
Otherwise, 1.1.52 Qi9›•Š « is applied and the ‚ 6T is in the place of 4, ( which brings ‚ 6T {æ by
1.1.50 v"4 1›t%T'5«.
167
Now, ‚ 6T !" is used in the next sūtra.
[*u*q `k ']( 6.1.125 6 ã" Q*S *4e' «( ~ #$| e" *‹T"3"' (

‚T#0|
When ‚ 6T or #0|ã is followed by QS, ( it remains as it is without *+ change.
1 7
‚ 6T or #0|ã QS (
1
#$| *Tn"u
( ã"5 1/3 Q*S 7/1 *4e' 0( « ~ #$| e" 3/1 *‹T"3"' 7/1
‚ 6T-#0| (

• ( ã"5 1/3 – ‚ 6T"š #0|ã"š ‚ 6T#0ã
‚ 6T-#0| | "5 (ID) «
• Q*S 7/1 – Though this can be taken from 6.1.77 p$9 3&*S«, it is again said in order to
show that the Q*S cannot cause any *+ change14; in ^% â'/.
• *4e' 0( – Always; this is to negate the possibility of 6.1.127 p$9› u&Ë 2"$ÞŠ •€š «.
• (
#$| e" 3/1 – #"*T^*o$ #$| *T, unmodified form; in T|T/3" by (u"þ) #$| e"*o8 ƒ^ û"4' «;
connected to nu*T, it means “stays as its own state”.
( ã"5 1/3) Q*S 7/1 #$| e" 3/1 Š56 III/3 «

[LSK] ‘T1 1/3 (= ‚ 6T-#0|
(
‚ 6T and #0|ã will remain as their own states, when QS follows.
There should be two types of examples. The first example is with ‚ 6T.
[LSK] w0• II/1 $| Žæ 8/1 Q` 0 0…5 1/1 S%*T III/1 - Come KIJ a--! Here a cow is moving.
Here, the *» of the first sentence (u"À) gets ‚ 6T by 8.2.84 %"}kT 1 S « ~ ‚ 6T5. Even when a
(
QS follows (
and 6.1.101 Q$5 u&Ë o/•5y « is applicable, the QS does not become a *4*'§, cause,
for any QS ( *+, by 6.1.125 ‚ 6T#0ã ( Thus the ‚ 6T remains as it is.
| " Q*S *4e' «.
Being w.3, a place of operation, of #$| *Tn"u, ‚ 6T should be regarded as * } even
though it is *`^"o/$"3y'. (
14
#$| *Tn"u is not $"3y. This is so that the Q*S does not become the cause for any QS-( *+.
| " Q*S *4e'«,( Q of Q` does not become
In the case of —"4 6 ƒ Q` «, ƒ is #0|ã by 1.1.14. By 6.1.125 ‚ 6T#0ã
*4*'§ for 3&-( *+. Since #$| *Tn"u is not $"3y on ƒ, u&yo/•y *+ is applied between —"4 6 and ƒ.
168
Qˆ *+5
| " Q*S *4e' «( is specifying two !"s:

The previous sūtra 6.1.125 ‚ 6T#0ã
1. ‚ 6T5 – defined by !" `k 1.2.27 ù$"i9› €o/•‚y 6T5« and enjoined by *u*q `k s in the section
starting from 8.2.82 u"ÀŠ »1 5 ‚ 6T ƒo"§5 « to 8.2.108. In this section of i•*6 }"t$…'o6 /, only
one sūtra 8.2.84 %"}kT 1 S « ~ ‚ 6T5 is introduced.
2. #0|ã5 – defined by !" `k s in the section starting from 1.1.11 o1o ( *ÒuS4 #0|ã' «( to 1.1.19.
In this section of i•*6 }"t$…'o6 /, four #0|ã !"-*uq"3$- `k s are introduced.
Now, we are going to see the first #0|ã !"-*uq"3$- `k .
[ !" `k ']( 1.1.11 o1o ( *ÒuS4 #0|ã' «(

The dual suffix ending with , ù, or ‘ is termed #0|ã.
( *ÒuS4' 1/1
o (-ùo (-‘T 1/1 ( #0|ã' 1/1
( «

• (
o (-ùo (-‘T 1/1 ( ùT S
– TS ( ‘T S
( (
o1T (SD)«; the T^%$%& is to make the sounds clearly
separated so that there is no ambiguity; this is adjective to *ÒuS4', ( thus Tot*u*q is
(
applied; “ , ù, ‘-ending dual suffix”; T becomes o ( by 8.2.39 ði" —29›t1« by being ô"t
in '" .
• (
*ÒuS4' 1/1 ( a !" given to *T™ ( and ^6 in
– *ÒuS4' is ( dual, defined by 1.4.102
6
T"Õ1$uS4*ÒuS4RºuS4"Õ1$25« and 1.4.103 ^5«; this is !/ for this sūtra.
• (
#0|ã' 1/1 – This is !".
[LSK] ( *ÒuS4' 1/1

o1o (-Qt' 1/1 ( #0|ã' 1/1
( Š"T III/1
( «
, ù, ‘-ending suffix in dual is termed #0|ã.
Note that this #0|ã !" is given to suffix.

We can expect three types of examples: in the case of , ù, and ‘-ending suffixes.
169
[LSK] ‹%/ 1/2 ‘T… 1/2 «

‹*% + ç 1/2, 2/2 (
^6 -#e3 with *ÒuS4- !" is suffixed.
‹% (+ 6.1.102 #‡'395 ^ku y u&y5 « ~ ‘$5 ^ku^y %395 Q$5 o/•5y
In the place of p and ç, ‘$"o12 is .
By 6.1.85 Qt"*ouí«, the ‘$"o12 is seen as *ÒuS4- !$ suffix. With the Zô1*2u "u (see
under 6.1.85 Qt"*ouí«), this single letter is seen as that which ends with . Thus, by 1.1.11
( #0|ã !" is given to this .
o1o ( *ÒuS4 #0|ã' «,
‹%/ + ‘T… Now, ‹%/ is followed by another word ‘T….
‹%/ ‘T… | " Q*S *4e' «( ~ #$| e"
6.1.125 ‚ 6T#0ã
6.1.77 p$9 3&*S « was negated.
[LSK] *uŽk 1/2 p'… 1/2 «

In exactly the same manner as above.
[LSK] 0¹1 1/2 Q'k 1/2 -

0¹" + ç 1/2, 2/2 (
^6 -#e3 with *ÒuS4- !" is suffixed.
0¹" + 2/ 7.1.18 ç™ w^5 « ~ 2/
(
ç is replaced by 2/ when preceded by ¥/#e3 w^-ending #"*T^*o$.
0¹" + ( pT and
By 1.3.8 i2òT*}T1 «, 2 is ( elided by 1.3.9 TŠ i9^5 «
0¹ +
( ‘ 6.1.87 w &5«
6 ~ ‘$5 ^ku^y %395 Q*S
6 (‘ by v"4›1 t%T'5).
In the place of w and , ‘$"o12 is 0&
By 6.1.85 Qt"*ouí«, the ‘$"o12 ‘ is seen as *ÒuS4- !$ suffix. With the Zô1*2u "u (see
under 6.1.85 Qt"*ouí«), this single letter ‘ is seen as that which ends with ‘. Thus, by
1.1.11 ( #0|ã !" is given to this ‘.
o1o ( *ÒuS4 #0|ã' «,
0¹1 + Q'k Now, 0¹1 is followed by another word Q'k.
0¹1 Q'k | " Q*S *4e' «( ~ #$| e"
6.1.125 ‚ 6T#0ã
6.1.78 ‘S9›3u"3"u5 « was negated.
170
Qˆ *+5
#0|ã- !"-*uq"3$- `k continues.
[ !" `k ']( 1.1.12 Qo 9 '"T «( ~ ( ã' (

T #0|
( the pronoun Qo are
and ù after ' of ( termed #0|ã.
( « ~ o (-ùT 1/1
Qo 5 6/1 '"T 5/1 ( #0|ã' 1/1
(

• Qo 5 6/1 – The pronoun “Qo ”;( in Qu3usé/ ( ¬+sé/) to '"T; ( “after ' of
( Qo , ( after ' as
(
a part of Qo ”.(
• '"T ( 5/1 – The letter “'”;( Q is for pronunciation; in *oÏ3901 ^Í'/ to ^%5, which can be
supplied to qualify T. (
• (
o (-ùT 1/1 ( ùT S
– TS ( (
T (SD) «; the T^%$%& is to make the sounds clearly separated
so that there is no ambiguity; this is !/.
• (
#0|ã' 1/1 – This is !".
( ^%… 1/2
[LSK] Q¼"T 5/1 T… 1/2 #0|ã… 1/2 Æ5 III/2 «
( a part of Qo ),( and ù are termed #0|ã.
After this (' as
Declension of the pronoun Qo (
^*• 6 ¹5 ¥/*i¹5 4^ 6 $*i¹5
‘$uS4' ( *ÒuS4' ( RºuS4' ( ‘$uS4' ( *ÒuS4' ( RºuS4' ( ‘$uS4' ( *ÒuS4' ( RºuS4' (
#‡'" Q … Q'k Q'/ Q • Q'k Q'k5 Qo5 Q'k Q'k*4
*ÒT/3" Q''6 ( Q'k Q'k4 ( Q'k' ( Q'k Q'k5 Qo5 Q'k Q'k*4
T|T/3" Q'46 " Q'k8"' ( Q'/*n5 Q'36 " Q'k8"' ( Q'k*n5 Q'46 " Q'k8"' ( Q'/*n5
ST‡6 * Q' 6 • Q'k8"' ( Q'/85 6 •

Q'³ Q'k8"' ( Q'k85 Q' 6 • Q'k8"' ( Q'/85
^Í'/ Q' 6 "T ( Q'k8"' ( Q'/85 6 "5

Q'³ Q'k8"' ( Q'k85 Q' 6 "T ( Q'k8"' ( Q'/85
sé/ 6
Q'³ Q'36 95 Q'/s"' ( Q'³
6 "5 Q'36 95 Q'ks"' ( 6
Q'³ Q'36 95 Q'/s"' (
â'/ Q'*6 4 ( Q'36 95 Q'/s 6 6 "' (

Q'³ Q'36 95 Q's6 6 Q'*6 4 ( Q'36 95 Q'/s 6
“*ÒuS4'”( from the previous sūtra is not taken as Q4u|6 *§ so that the RºuS4 of “Q'/” in
masculine gets #0|ã !".
171
( the declined form of Qo are

Both and ù, as well as the ' of ( $"3y of sūtras in *`^"o/.
Namely, 8.2.80 Qo 9› o1 "yÊ o9 '5« and 8.2.81 ‘T o ( RºuS4 1«. 1.1.12 Qo 9 '"T «( can recognize
these $"3ys because of “uS4- "'Wy'”,( capacity of the sūtra.
Example of RºuS4 –
[LSK] Q'/ 1/3 2"5 1/3«
( — (
Qo +
( 2/
Qo + 7.1.17 — 5 2/ « ~ uy4"Ä5
Qo Q + 7.2.102 eo"o/4"'5 « ~ *un´…
Qo + 6 « ~ ^%–^' (
6.1.97 QT9 0&1
Note that Q + Q in the #"*T^*o$ is Qt%¹, having precedence, to Q + between
(
#"*T^*o$ and ^6 -#e3, which is R*‹%¹.
Qo1 6.1.87 w &5
6 « ~ Q*S
Q'/ ( '5
8.2.81 ‘T o ( RºuS4 1 « ~ Qo 5 o"T o5
( Qo gains
The after ' of ( #0|ã- !" by 1.1.12 Qo 9 '"T «( ~ ( ã' (
T #0|
Q'/ + 25
Q'/ 25 | " Q*S *4e' «(
6.1.125 ‚ 6T#0ã
Example of *ÒuS4 ù –
[LSK] %"'$| Ž… 1/2 Q'k 1/2 w "T1 II/2« These two, Rāma and KIJ a, sit.
( ç
Qo +
Qo Q + ç 7.2.102 eo"o/4"'5 « ~ *un´…
Qo + ç 6 « ~ ^%–^' (
6.1.97 QT9 0&1
(
Note that Q + Q in the #"*T^*o$ is Qt%¹ to Q + ç between #"*T^*o$ and ^6 -#e3,
which is R*‹%¹.
Qo… 6.1.88 u|*}%1*S « ~ wT (
Q'k ( o5 '5 «
8.2.80 Qo 5 Q 51 o"T ƒ
( Qo gains
The ù after ' of ( #0|ã- !" by 1.1.12 Qo 9 '"T «( ~ ( ã' (
T #0|
Q'k + w "T1
Q'k w "T1 | " Q*S *4e' «(
6.1.125 ‚ 6T#0ã
( done.
Now, ô$| e for the word '"T is
172
Qˆ *+5
( ?( Q'$
[LSK] '"T *$' 6 1 ›` -
( Q (
Qo ( + Q$• S + ( ( »1 5«
5.3.71 QZ3 uy4"Ä"'$• S #"$
( suffixed before *», in the sense of itself (€"‡Ë).
T*}T-#e3 Q$• S is
Qo$ ( 6
Q4R+i9^5 and u&y ?1i4' (
6
Q'$ ( ( o5 '5 «
8.2.80 Qo 5 Q 51 o"T ƒ
( — (
6 +
Q'$
Q'$ ( 2/
6 + 7.1.17 — 5 2/ « ~ uy4"Ä5
6 Q+
Q'$ 7.2.102 eo"o/4"'5 « ~ *un´…
6 +
Q'$ 6 « ~ ^%–^' (
6.1.97 QT9 0&1
(
Note that Q + Q in the #"*T^*o$ is Qt%¹ to Q + between #"*T^*o$ and ^6 -#e3,
which is R*‹%¹.
6 1
Q'$ 6.1.87 w &5
6 « ~ Q*S
( Qo , ( thus ‘ does not get #0|ã- !" by 1.1.12 Qo 9 '"T «( ~
The ‘ is not after ' of T. (
6 1 + Q`
Q'$
6 1 ›`
Q'$ 6.1.109 ‘™5 ô"t"o*T « ~ ^ku5y
Side note:
Even though and ù of Q'k8"', ( Q'/*n5, etc., satisfy the conditions for getting #0|ã !" told
in 1.1.12 Qo 9 '"T«,( since there are letters after the and ù, and hence 6.1.125 ‚ 6T#0ã
| " Q*S
*4e' «( is not applicable, there is no use for giving #0|ã !". [Bh]
173
Now we are seeing two *4^"T- !"-*uq"3$- `k s because the next #0|ã !"-*uq"3$- `k
requires *4^"T- !".
[ !" `k ']( 1.4.57 S"o39› 1 « ~ *4^"T"5

Words in S-w*o-0& (a group of words beginning with S), in the sense of Q (that
which does not have gender or number), gain *4^"T !".
S-wo35 1/3 Q 1 7/1 « ~ *4^"T"5 1/3

• ( T1 1/3 S"o35 1/3 (116B)«; a group of words
S-wo35 1/3 – S-2²5 1/1 w*o5 1/1 31s"' 6/3
beginning with S; this is !/.
• Q 1 7/1 – means LZ, that which has gender and number; 4 ( •', ( that
'Q
which does not have gender or number; in *us3 â'/.
• *4^"T"5 1/3 – This is !".
[LSK] QLZ"‡"y5 1/3 S"o35 1/3 *4^"T"5 1/3 Š56 III/3 -

Words in S"*o0&, whose meanings are without gender or number are termed *4^"T.
The members of S"*o0& are found in QZ3"*4 section in i•*6 }"t$…'o6 /.
174
Qˆ *+5
[ !" `k ']( 1.4.58 #"o35 « ~ Q 1 *4^"T"5

Words in #-w*o-0& (a group of words beginning with #), in the sense of Q (that
which does not have gender or number), gain *4^"T !".
#-wo35 1/3 « ~ Q 1 7/1 *4^"T"5 1/3

• ( T1 1/3 #"o35 1/3 (116B)«; a group of words
#-wo35 1/3 – #-2²5 1/1 w*o5 1/1 31s"' 6/3
beginning with #; this is !/.
• Q 1 7/1 – means LZ, that which has gender and number; 4 ( •', ( that
'Q
which does not have gender or number; in *us3 â'/.
• *4^"T"5 1/3 – This is !".
[LSK] ‘T1 1/3 Q*^ 0 T‡" 0 (*4^"T"5 1/3 Š56 III/3) -

Words in #"*o0&, whose meanings are without gender or number are termed *4^"T.
The members of #"*o0& are found under 1.4.59 ƒ^ 0"y5 *¨3"3901 in i•*6 }"t$…'o6 /.
1. # 2. ^%" 3. Q^ 4. '( 5. Q4 6 6. Qu
7. *4 ( 8. *4% ( 9. Ê ( 10. Ê% ( 11. *u 12. w™ (
13. *4 14. Q*q 15. Q*^ 16. Q*T 17. 6 18. ƒo (
19. Q*n 20. #*T 21. ^*% 22. ƒ^
Summary of !"s to #"*o0&
!/ !" !"#o125
1.4.58 #"o35 *4^"T"5 1.1.37 €%"*o*4^"T'Z3' «(
1.4.59 #"o35 *¨3"3901 ƒ^ 0"y5 8.4.14 ƒ^ 0"yo '" ›1 *^ &9ô12Š « ~ &5 45
1.4.60 #"o35 *¨3"3901 0*T5 2.2.18 $60*T#"o35 « ~ T°7s5

6 '" 5
1.4.61 ~ 79 *X, etc…
175
Now the *4^"T- !" is used in the next #0|ã !"-*uq"3$- `k .
[ !" `k ']( 1.1.14 *4^"T ‘$"—4"™ ( « ~ #0|ã' (

*4^"T, which is a single vowel, but not w™,( is termed #0|ã.
( Q4"™ ( 1/1 « ~ #0|ã' 1/1

*4^"T5 1/1 ‘$-QS 1/1 (

• *4^"T5 1/1 – That which is termed *4^"T by *4^"T !"-*uq"3$ `k s from 1.4.56 to 1.4.98; this
is adjective to ‘$"S. (
• (
‘$-QS 1/1 (
– ‘$5 S Q … QS ‘$"S (
(KT); this is !/.
(
Here ‘$"S does not mean “a group of letters consisting of only one vowel, and the rest
(
are consonants”. This ‘$"S means “a single vowel”.
• Q4"™ ( 1/1 – 4 w™ ( Q4"™ ( (NT); this is adjective to ‘$"S. (
(
Note that conditions regarding letters are always after removing pT letters.
• (
#0|ã' 1/1 – This is !".
( *4^"T5 1/1 w™u( —y5 1/1 #0|ã5 1/1 Š"T III/1

[LSK] ‘$5 1/1 QS 1/1 ( «
The single vowel which is *4^"T, other than w™,( is termed #0|ã.
(
Examples for ‘$"S *4^"T are:
[LSK] p pŒ5 « ƒ ƒ'125 «
Both p and ƒ are found in S"*o0&, and their meanings are *u¼3, wonder, and *uT$y ,
doubt, both of which are Q . Thus #0|ã !" is given by 1.1.14 *4^"T ‘$"—4"™«( . Then even
(
when QS follows, | " Q*S *4e' «,
by 6.1.125 ‚ 6T#0ã (
( the following QS does not become the
*4*'§, cause for QS-( *+, here u&yo/• y *+.
176
Qˆ *+5
Now, Q4"™ ( is explained.

There are two types of ws:
• (
w™ ( which is called *™T (that which is with ™ ( as pT).(
• (
w which is called Q*™T (that which is without ™ ( as pT).(
In usage they look the same. However, the distinction is made because there is a
difference in *+. When w is Q*™T, ( it gains #0|ã !" by this sūtra 1.1.14 *4^"T ‘$"—4"™«( and
(
| " Q*S *4e' «.
no *+ happens even when a vowel follows, by 6.1.125 ‚ 6T#0ã
(
The meanings of *™T w™ ( are told in the following śloka:
( and Q*™T w
2o‡Ë *¨3"3901 '3"yo"*n*uq… S 35 «

(
‘T'"T *™T *u±"o ( u"À¼%&39%*™T -
• w™ ( (*™T)(
1. (
sT-Q‡Ë – in the sense of “slight, little” – e.g., w(™)( ( sT)( ƒŽ' ,Ž'
( «(
2. *¨3"3901 – ƒ^ 0y5 – e.g., w(™)( + p*‹ = ‘*‹ «
6 15
3. '3"yo" – exclusive limit – e.g., w(™)( w '´ "%5 «
6
4. Q*n*u*q5 – inclusive limit – e.g., w(™)( ÚÛnu4"T «(
• w (Q*™T)(
1. u"À1 – but now – e.g., w ‘u 'Õ 1 « But now, you think that way. (Before you were
not thinking that way, but now you are.)
2. ¼%&1 – recollection – e.g., w ‘u *$i TT «( Ah, I remember it was indeed that.
Taking this into account, the following examples show how Q4"™ ( in the sūtra works.
[LSK] “u"À¼%&39%*™T”;( w ‘u 4 6 'Õ 1 « w ‘u *$i TT «( QÕ` *™T; ( w sÊŽ' ,Ž'
( (
-
(
By the statement “Q*™T (w) is in the sense of u"À and ¼%&”, these ws in the
following sentences are considered to be Q4"™ ( and thus #0|ã !" is given, resulting in non-
(
application of *+. In other senses, ws are considered to be *™T (w™)( , thus #0|ã !" is not
given, resulting in application of *+ change.
177
[ !" `k ']( 1.1.15 ,T «( ~ *4^"T5 #0|ã' (

,$"%-ending *4^"T is termed #0|ã.
( « ~ *4^"T5 1/1 #0|ã' 1/1

,T 1/1 (
1 word in the `k , 2 words as Q4u|6 *§.

• (
,T 1/1 ( limit to o/•;y this is adjective to *4^"T5; being adjective, Tot*u*q is
– , with T to
applied, resulting in “,$"%-ending *4^"T”.
• *4^"T5 1/1 – That which is termed *4^"T by *4^"T !"-*uq"3$ `k s from 1.4.56 to 1.4.98.
• (
#0|ã' 1/1 – This is !".
(
[LSK] ,ot5 1/1 *4^"T5 1/1 #0|ã5 1/1 Š"T III/1 «
The ,$"%-ending *4^"T is termed #0|ã.
[LSK] Q‹9 0 2"5 8/3 -

Q‹9 is termed *4^"T because it is in S"*o0& and it is Q u*S4. ( Thus Q‹9 is an ,$"%-
( #0|ã !" is gained. Then by 6.1.125 ‚ 6T#0ã
ending *4^"T. By 1.1.15 ,T «, (
| " Q*S *4e' «, of
2"5, though being a vowel, does not become a *4*'§ for 6.1.78 ‘S9›3u"3"u5 «.
Other ,-ending *4^"Ts are: Q‡9 (now), 49 (no), w‹9 (interjection expressing doubt or
alternative), ƒT"‹9 (questioning).
178
Qˆ *+5
[ !" `k ']( 1.1.16 6 2"$ÞŠ1T"u4"sË « ~ ,T #0|

¬}… ( ã' (
,$"% caused by ¬*6 } in Qu•*o$ literature, when followed by p*T, is termed #0|ã in
the opinion of 2"$Þ '*6 4.
( #0|ã' 1/1
¬}6 … 7/1 2"$ÞŠ 6/1 pT… 7/1 Q4"sË 7/1 « ~ ,T 1/1 (

• ¬}6 … 7/1 – (
¬*6 } is !" for ,• 6 singular in ¬9q4 of ^6 -#e3, by !" `k 2.3.49 ‘$uS4
¬*6 }5« ~ #‡'" ¬9q4 1; in *4*'§ â'/.
• 2"$ÞŠ 6/1 – The name of a grammarian respected by ^"*&*4, in ¬+sé/ to 'T1 (in the
opinion). This implies that this *u*q is optional.
• pT… 7/1 – #"*T^*o$ is p*T; in ^% â'/.
• ( 5 « that which is connected to x*s; 4 wsy5 Q4"sy5 (NT); in
Q4"sË 7/1 – xs15 po' wsy
*us3 â'/.
• ,T (1/1 – From 1.1.15 ,T «;
( , with T to
( limit to o/•.y
• (
#0|ã' 1/1 – This is !".
[LSK] ¬*6 }-*4*'§$5 1/1 ,$"%5 1/1 u" 0 #0|ã5 1/1 Qu•*o$1 7/1 pT… 7/1 ^%1 7/1 «
,$"% which is caused by ¬*6 } ( ¬9q4 singular), when followed by p*T, is optionally
#0|ã in non-Vedic literature.
The following is the preparation of the example given in the u|*§.

*uŽ 6 + • 6 ( ¬*6 }5) • 6 is enjoined by 2.3.47 ¬9q4 1 S « ~ #‡'"
This • 6 is termed ¬*6 }5 by 2.3.49 ‘$uS4 ¬*6 }5« ~ #‡'" ¬9q4 1
*uŽ9 + ( 6 « ~ ¬}6 …
7.3.108 •€Š 0&5
*uŽ9 6.1.69 ‘™-( •€"T ( ¬}6 51 « ~ i9^5
This , of *uŽ9 is caused by the suffix ¬*6 }.
*uŽ9 + p*T When p*T follows this , in non-Vedic literature, #0|ã !" is
optionally given to the , by 1.1.16 ¬}6 … 2"$ÞŠ1T"u4"sË «.
179
There are three forms possible.

[LSK] *uŽ9 p*T, *uŽ p*T, *uŽ*u*T -
Case 1) Optional #0|ã !"

*uŽ9 + p*T The , of *uŽ9 caused by the suffix ¬*6 } is termed #0|ã optionally
by 1.1.16 ¬}6 … 2"$ÞŠ1T"u4"sË «.
*uŽ9 + p*T | " Q*S *4e' «( prevents any Q*S- *+ to happen.
6.1.125 ‚ 6T#0ã
(
Case 2) No #0|ã !" with optional u-i9^
*uŽ9 + p*T
( +
*uŽ Qu ( p*T 6.1.78 ‘S9›3u"3"u5 « ~ Q*S
( + p*T
*uŽ Q y Q*2
8.3.19 i9^5 2"$ÞŠ « ~ ôŠ Z95 Q^kuŠ
*uŽ p*T
(
Case 3) No #0|ã !" and no u-i9^
*uŽ9 + p*T
( +
*uŽ Qu ( p*T 6.1.78 ‘S9›3u"3"u5 « ~ Q*S
*uŽ*u*T
180
Qˆ *+5
The next sūtra is an optional Qû"o to #0|ã !" by 1.1.14 *4^"T ‘$"—4"™«( .
[ !" `k ']( 8.3.33 '3 ƒÈ9 u9 u" « ~ Q*S

(
For the particle ƒÈ, ( when preceded by '3 and followed by QS, ( u is
( the substitute
optionally. 5 6 7
'3 ( ƒÈ ( QS (
1
Optional u (
'35 5/1 ƒÈ5 6/1 u5 1/1 u" 0 « ~ Q*S 7/1

• '35 5/1 – #e"‹"%5 '3; ( in ^ku^y Í'/.
• ƒÈ5 6/1 – A particle ƒÈ15( in S"*o0&; since it is Q , *4^"T !" is given by 1.4.57
S"o39› «1 ; #0|ã !" is possible by 1.1.14 *4^"T ‘$"—4"™ ( «; in v"431 90" sé/.
• ( for pronunciation (ƒí"%&"‡Ë); this is wo12.
u5 1/1 – u$"%5; Q after u is
• u" 0 – This operation is optional.
• Q*S 7/1 – #e"‹"%5 QS; ( in ^% â'/.
[LSK] '35 5/1 ^%Š 6/1 ƒÈ5 6/1 u5 1/1 u" 0 Q*S 7/1 «
( the substitute optionally in the place of ƒÈ which
u is ( (
is after '3 when (
QS follows.
Two forms are possible.
6 ', ( *$' 6 ƒ´' -
[LSK] *$Y´ (
Case 1) u$"%-wo12 option

( )( ƒ´' (
*$' ƒ(È
(
*$' u( ƒ´' ( 8.3.33 '3 ƒÈ9 u9 u" « ~ Q*S
6 '(
*$Y´ ( Q* } for 8.3.23 '9›4€"%5
The u is 6 6
«, hence Q4€"% does not occur.
Case 2) #0|ã !"
( )( ƒ´' (
*$' ƒ(È ƒ is #0|ã by 1.1. 14 *4^"T ‘$"—4"™ ( « ~ #0|ã' (
( ƒ´' (
*$' ƒ | " Q*S *4e' «( ~ #$| e"
6.1.125 ‚ 6T#0ã
*$' 6 ƒ´' (
15
ƒ Quq"%&1 « QŠu• *¨3"*u21s"‡Ñ *ÈB%&' «( [AK]
181
#0|ã topic is over now. The next sūtra is another rule of #$| *Tn"u. There are two
things told in this one sūtra.
[*u*q `k ']( 6.1.127 p$9› u&Ë 2"$ÞŠ •€š « ~ ô"tŠ Q*S #$| e" *‹T"3"' (
When p$ ( is followed by Q u&y vowel, there is #$| *Tn"u (remaining without *+) and
the p$ ( becomes •€.
6 7
ô p$ ( Q u&y QS ( 1
#$| *Tn"u
1
•€ (no *+)
(
p$5 6/1 Q u&Ë 7/1 2"$ÞŠ 6/1 •€5 1/1 S 0 « ~ ô"tŠ 6/1 Q*S 7/1 #$| e" 3/1 *‹T"3"' 7/1
• p$5 6/1 – #e"‹"%5 p$;( in v"431 90" sé/.
• ( u&y' (NT),
Q u&Ë 7/1 – 4 u&y' Q ( ( this is an adjective to Q*S; in ^% â'/.
T*¼4 «;
opinion).
• •€5 1/1 – This is wo12; ^*%n"s" `k 1.2.28 QSš « brings QS5 6/1 in v"431 90" sé/, resulting in
“in the place of QS”( to be added in u|*§.
• S 0 – There are two injunctions in this sūtra. One is to keep it without *+, another is
to make p$ ( short.
• ô"tŠ 6/1 – From 6.1.109 ‘™5 ô"t"o*T« by 'Á¸k$‚ 6T0*T5; *un*´*u^*%&"' has happened;
this is adjective to p$5.
• Q*S 7/1 – From 6.1.77 p$9 3&*S«; in ^% â'/.
• (
#$| e" 3/1 – #"*T^*o$ #$| *T, unmodified form; in T|T/3" by (u"þ) #$| e"*o8 ƒ^ û"4' «.
[LSK] ô"t"5 1/3 p$5 1/3 •€"5 1/3 u" 0 Š56 III/3 Q u&Ë 7/1 Q*S 7/1 « 16
(
p$s( at the end of ô become short optionally when Q u&y QS follows.
16
( ‘$
[SK] ô"t"5 p$5 6/1 Q u&Ë Q*S ^%1 #$| e"5 Š56 •€5 S u" « [B] ô"tŠ p$5 Q u&Ë Q*S ^%1 •€5 Š"T p*T
u"À' «( •€5 #$| e" €n"u14 Qu*TéT1 p*T u"À"t% †±T1 « This reading seems better.
182
Qˆ *+5
Note that i•*6 }"t$…'o6 /$"% takes all 1st case to show #$| *Tn"u is not a change.
( 4 0 €%- *+5 1/1 «

[LSK] •€-*u*q- "'W"yT 5/1
Because of the effect of •€, no vowel *+ is applied after making •€, otherwise,
the statement would become useless.
[LSK] S*¨ Q`, SZ` «

Two forms are possible.
Case 1)
S¨/17 + Q`
S*¨ Q` 6.1.127 p$9› u&Ë 2"$ÞŠ •€š «
Case 2)
S¨/ + Q`
( Q`
SZ + 6.1.77 p$9 3&*S«
SZ`
[LSK] ô"t"5 p*T *$'?( 0"[> –

Why was “ô"t"5” told? To achieve the form “0…[>”, which requires the next sūtra –
17
(
8.2.7 4i9^5 #"*T^*o$"tŠ « is not Q* } for the sūtra in ^"o â"j"3/ because of 8.2.2 4i9^5 ^6 -€%-
(
!"-T*6 Ïu*qs 6 $| *T «, which limits the scope of Q* } of 4i9^ to those sūtras for ^6 -*u*q, €%-*u*q, !"-*u*q,
(
and T*6 Ïu*q when $| T follows. For *+-*u*qs, 4i9^ is * }.
183
The next sūtra is about ‹i-( *+. To achieve the form mentioned in the last sūtra, we
need to know this sūtra.
[*u*q `k ']( 8.4.46 QS9 %‹"8" Ò1 « ~ 3%5 u"

There is optional doubling of 3% (when it is preceded by % (or ‹ ( which is preceded by
QS. (
5 5 6
QS ( % (or ‹ ( 3% (
1
*Ò•
( Ò1 1/1 « ~ 3%5 6/1 u" 0

QS5 5/1 %‹"8"' 5/2
• QS5 5/1 – #e"‹"%5 QS; ( in ^ku^y Í'/.
• (
%‹"8"' 5/2 – %5 S ‹5 S %‹… (ID) T"8"'; ( Q after % (and ‹ ( are for pronunciation; in ^ku^y Í'/.
• Ò1 1/2 – #"*T^*o$ *Ò; this is wo12; the letter becomes double.
• 3%5 6/1 – #e"‹"%5 3%;( in v"431 90" sé/.
• u" 0 – This is optional.
( %1Ý‹$"%"8"' 5/2
[LSK] QS5 5/1 ^%"8"' 5/2 ( ^%Š 6/1 3%5 6/1 Ò1 1/2 u" 0 Æ5 III/2 «
There is optional doubling of 3% (after %1Ý or ‹ ( , which is after QS. (

[LSK] 0…[> «
0…%/ + ç #"*T^*o$ 0…%/ ( $"%"t-¥/*i¹-2²5) is declined in 1/2 or 2/2.
( ç
0…% (3 + 6.1.77 p$9 3&*S «
( +
0…% (3 3 ( ç 8.4.46 QS9 %‹"8" Ò1 « ~ u" 3%5
0…[>
When *Ò• is not taken, “0…3>” is another form.
( '"4', (
Other examples for the doubling by 8.4.46 QS9 %‹"8" Ò1 « are: u§y'"4'/uTy
$/*§y5/$/*Ty5, etc.
The next u"*Ty$ is *4s1q (prohibition) of 6.1.127 p$9› u&Ë 2"$Þ •€š «.
184
Qˆ *+5
(u"*Ty$')( 4 '" 1 «
#$| *Tn"u and •€ by 6.1.127 p$9› u&Ë 2"$Þ •€š « do not happen when two words
are compounded.
4 0 '" 1 7/1 «
2 words in the `k , other words are understood by the context.
• 4 0 – Prohibition to the two effects of 6.1.127 p$9› u&Ë 2"$Þ •€š «, namely, #$| *Tn"u
and •€, even when all the conditions are met.
• '" 1 7/1 – Compound, as per '" *u*q starting from 2.1.1 to 2.2.38; in Q*q$%& â'/.
[LSK] u"AÂ5 -
(
u"A"' QÂ5 «
6 « ~ ^6 ( '" 5
u"^/ + *™ + QÂ + • 6 « 2.1.4 ‹ ^"
#"*T^*o$ !" by 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
u"^/ + QÂ 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
The ô"t followed by Q u&y Q is satisfying all the conditions of
6.1.127 p$9› u&Ë 2"$Þ •€š «. However, this is negated by
(u"þ) 4 '" 1 «
( QÂ
u"^ 3( + 6.1.77 p$9 3&*S «
u"AÂ
185
[*u*q `k ']( 6.1.128 xe$5 « ~ 2"$ÞŠ •€5 ô"tŠ #$| e" *‹T"3"' (
In the opinion of 2"$Þ, when Q$ ( is followed by short x, there is #$| *Tn"u
(remaining without *+) and the Q$ ( becomes •€.
6 7
Q$ ( xT ( 1
#$| *Tn"u
1
•€ (no *+)
(
x*T 7/1 Q$5 6/1 « 2"$ÞŠ 6/1 •€5 1/1 ô"tŠ 6/1 #$| e" 3/1 *‹T"3"' 7/1
• x*T 7/1 – Short x; in ^% â'/.
• Q$5 6/1 – #e"‹"%5 Q$;( in v"431 90" sé/.
opinion).
• •€5 1/1 – This is wo12; ^*%n"s" `k 1.2.28 QSš « brings QS5 6/1 in v"431 90" sé/, resulting
in “in the place of QS”( to be added in u|*§.
• ô"tŠ 6/1 – From 6.1.109 ‘™5 ô"t"o*T« by 'Á¸k$‚ 6T0*T5; *un*´*u^*%&"' has happened;
this is adjective to Q$5.
• (
#$| e" 3/1 – #"*T^*o$ #$| *T, unmodified form; in T|T/3" by (u"þ) #$| e"*o8 ƒ^ û"4' «.
[LSK] x*T 7/1 ^%1 7/1 ô"t"5 1/3 Q$5 1/3 #"ÏuT 0( (•€"5 1/3 #$| e" 3/1) u" 0 Š56 III/3 «
When short x follows, Q$ ( at the end of ô does not undergo *+ change and
becomes •€.
[LSK] ÚÛ x*s5 « ÚÛ*sy5 «
ÚÛ" 1/1 + x*s5 1/1
6.1.78 w &5«
6 is #"â, but negated by
ÚÛ x*s5 6.1.128 xe$5 « ~ ô"tŠ #$| e" •€5 2"$ÞŠ
If this option is not taken:
ÚÛ*sy5 6.1.78 w &5«
6
186
Qˆ *+5
ô$| e for the word ô"t"5 is done.

[LSK] ô"t"5 *$'?( w•yT -
(
If the Q$ ( is not at the end of ô, 6.1.128 xe$5« does not apply. In this
counterexample, w» (-w0' is used to illustrate Qô"t Q$.(
w0' is not an independent entity. It is a part of what it is attached to. Thus, w0'
does not get ô status. In this example, w» (-w0' is a part of q"T.6
( $Ty*%/III/1
x (1P) to go + i™/
x + i™ ( 3.2.111 Q4±T4 1 i™ ( « ~ nkT 1
x + *T 3.4.78 *Tâ*÷ð… « ~ iŠ
x+T( 3.4.100 pTš « ~ i9^5
( T(
x + 2^ + 3.1.68 $Ty*% 2^ «( ~ "uyq"T$
6 1
x• ( + Q + T ( 7.3.78 ^"\":"v"Ä"o"Á”2*Ty'*Ty2o o" *^R*—\q'*Té'43•-^Ÿ3•yq…2/3 /o"5« ~

*2*T
w» ( + x• ( + Q + T ( 6.4.72 w¸—"o/4"' «( ~ ( 6
i]i|™“
w% (+ • ( + Q + T ( 6.1.90 w»š « ~ u|*}5
Note: q"T 6 can also be x• ( (6P) to go, to make the same result.
( #$| *Tn"u5 1/1 « âxs/&"' –( âs*&"' «(

Side note: [SK] '" 1 7/1 Q*^ 0 Q3' 1/1
The u"*Ty$ 4 '" 1 « for the previous sūtra does not apply for this sūtra.
[LSK] p*T QS-( *+5 «

Thus ends QS-( *+18 section.
18
( QS should
S of ( (
have been subject to 8.2.39 ði" —29›t1«, 8.4.53 ði" —2 ð*2«, and 8.4.55 Ó*% S« to
( retained without *+.
become $.( However, to avoid confusion with #e"‹"% Q$,( QS is
187
Summary of QS ( *+5
These are the Qˆ *+ sūtras studied in QS ( *+5 section in i•*6 }"t$…'o6 /. They can be
divided into three sections.
( *+-#$%&')(
6.1.72 *‹T"3"' ( 158 QS-( *+ section starts from here.
6.1.77 p$9 3&*S Q*S 125
6.1.78 ‘S9›3u"3"u5 ‘S5 83
6.1.79 u"t9 *3 #e31
(‘$"o12"*o-#$%&')(
6.1.84 ‘$5 ^ku^y %395 111 ‘$"o12 section starts from here.
6.1.85 Qt"*ouí All ‘$"o12s are Qt-w*ouT (
6
6.1.87 wo ( 0&5 (
wT 96 6
0&-wo1
2
6.1.88 u|*}%1*S u|*}5 92; ‘*S 89 u|*}-wo12
6.1.89 ‘e1qek’ 6
6.1.91 ƒ^ 0"yo ( x*T q"T… ƒ^ 0"yo (, q"T… 94; x*T 92
6.1.94 ‘*™ ^%–^' ( (
^%–^' 100 ^%–^-wo12
6.1.95 ,'"™9š
6.1.101 Q$5 u&Ë o/•5y Q$5 107; o/•5y 106 o/•-y wo12
6.1.109 ‘™5 ô"t"o*T ‘™5 110; Q*T 113 y ^-wo12
^ku–
(#$| *Tn"u"*o-#$%&')(
6.1.122 uy` *un"s" 095 *un"s" 123; 095 124
6.1.123 Qu™ ( ê9»"34Š Qu™ ( 124
6.1.124 pŒ1 S
| " Q*S *4e' (
6.1.125 ‚ 6T-#0ã Q*S 130 Some #0|ã !"s are studied.
6.1.127 p$9› u&Ë 2"$ÞŠ •€š
2"$ÞŠ •€5 S 128
6.1.128 xe$5
188
Qˆ *+5
Understanding how the Pā ini system works
In the beginning of QS-( *+ section, the basic format of the diagram is like this:
6 7
v"4/ ^%*4*'§5
v"4' (
1
wo125
In QU"j"3/ `k ^"l5, one can easily understand the meaning of a sūtra by bringing
down Q4u|6 *§ and identifying the case of each word.
In the following chart, each word is sorted by case and Q4u|6 *§s are shown for easy
understanding.
type of sūtra wo125 ($"3y')( v"4/ ^%*4*'§5 *us35

st th th
sūtra number in 1 case in 6 case in 7 case (topic)
Q*q$"%5 6.1.72 - - - *‹T"3"' (
*u*q5 6.1.77 3& ( p$5 Q*S
*u*q5 6.1.78 Q3u"3"u5 ‘S5
*u*q5 6.1.79 u"t5 (Qu, ( wu)( (,, ç) *3 #e31
Q4u|6 *§ is not taken in a sūtra in which a word is already told in the same case and
sense as the Q4u|6 *§ word. For example, in 6.1.77 p$9 3&*S«, Q*S in ^% â'/ is told to be Q4u|6 *§
up to the sūtra number 6.1.125. However, in 6.1.79 u"t9 *3 #e31 «, *3 and #e31 are told in
^% â'/. In this case, bringing Q*S does not serve any purpose. Thus Q4u|6 *§ is skipped.
189
In the Q*q$"% (topic) of ‘$5 ^ku^y %395, the basic format of the diagram is like this:
5 7
^ku5y ^%5
v"4' ( 6
1
‘$5 wo125
Observe well the importance of `k ¨', the order of sūtra.
type of sūtra wo125 ($"3y')( ^ku5y ^%5 *us35

st th th
sūtra number in 1 case in 5 case in 7 case (topic)
Q*q$"%5 6.1.72 *‹T"3"' (
*u*q5 6.1.77 3& ( p$5 Q*S
Q*q$"%5 6.1.84 ‘$5
Q*To125 6.1.85 Qt"*ouT (
*u*q5 6.1.87 6
0&5 ( T)(
wT (Qu&"y
*u*q5 6.1.88 u|*}5 ‘*S
*u*q5 6.1.89 ‘*T-‘q*T-ù’ 6
*u*q5 6.1.91 ƒ^ 0"yT ( x*T q"T…
*u*q5 6.1.94 ^%–^' ( ‘*™
*u*q5 6.1.95 (
,'-w™95
*u*q5 6.1.101 o/•5y Q$5 u&Ë
*u*q5 6.1.109 ^ku5y ô"t"T, ( ‘™5 Q*T
Small questions:
1. How many types of ‘$"o12 are there? What are they?
190
Qˆ *+5
Summary of !" `k s studied in QS-( *+ section
!" `k ' ( !" !"#o125

1.1.1 u|*}%"o•S «( u|*}5 6.1.88 u|*}%1*S«, 7.2.114 '|—u1 *| y }5«
6
1.1.2 Qo1™ ( 0&5« 6
0&5 6.1.87 w &5«,
6 y6 5«, 6.4.146 ,0&
7.2.82 *'o10& y6 5«
1.1.11 o1*ÒuS4 #0|ã' «( #0|ã' ( | " Q*S *4e'«(
6.1.125 ‚ 6T#0ã
1.1.64 QS9›•"*o *» « *» 3.4.79 *»T wö4 1ô"4" »1 %«1 , 6.4.143 »1 5«
1.1.70 T^%ÆB"iŠ« T^%5 Q& ( 6
1.1.2 Qo1™ ( 0&5«, 1.1.1 u|*}%"o•S«,( 6.1.91 ƒ^ 0"y”*T q"T…«
1.3.1 nku"o39 q"Tu5 « q"T56 6.1.91 ƒ^ 0"y”*T q"T…«, 3.1.91 q"T95«
1.4.57 S"o39› 1« *4^"T5 1.1.14 *4^"T ‘$"—4"™«( 1.1.37 €%"*o*4^"T'Z3'«(
1.4.58 #"o35 « *4^"T5 1.1.14 *4^"T ‘$"—4"™«( 1.1.37 €%"*o*4^"T'Z3'«(
1.4.59 ƒ^ 0"y5 *¨3"3901 « ƒ^ 0y5 6.1.91 ƒ^ 0"y”*T q"T…«, 8.4.14 ƒ^ 0"yo " ›1 *^ &9ô12Š«
Summary of *`^"o/ `k s studied in QS-( *+ section
8.2.1 ^ku`y * }' «( Q*q$"% `k ' (

8.2.23 390"tŠ i9^5« ~ ôŠ ^ku`y * }' (
8.2.84 %"}kT 1 S « ~ ‚ 6T5 ôŠ ^ku`y * }' (
8.3.19 i9^5 2"$ÞŠ « ~ ôŠ ^ku`y * }' ( *u*q `k "*&
8.4.46 QS9 %‹"8" Ò1 « ~ 3%5 u" ^ku`y * }' (
8.4.47 Q4*S S « ~ QS5 3%5 u" ^ku`y * }' (
(
8.4.53 ði" —2 ð*2 « ~ ^ku`y * }' (
Summary of ^*%n"s" `k s studied in QS-( *+ section
See the ^*%n"s" `k section in the Introduction.
191
Q‡ ‹i-( *+5
Now, the section of *+ rules between consonants starts.
‹i-( *+5 1) š•/U

6 •6 - *+5 “%"'Ÿ21T”1
[*u*q `k ']( 8.4.40 6 š56 « ~ *‹T"3"' (

Æ95 š4"
(
2 and Su0y are the wo12s for (
and Tu0y, respectively, when they meet.
3 6 6 3
( 6
2/S ( 6
/T ( 6
/T ( 6
2/S
1 1
( 6
2/S ( 6
2/S
Æ95 6/1 š4" (

6 3/1 š56 1/1 « ~ *‹T"3"' 7/1

• Æ95 6/1 – #"*T^*o$ is Æ;6 (
and Tu0y; ( T56 S Æ56 (SD), TŠ «; in v"4 1390" sé/ (1.1.49 sé/
S
( thus it is !" for 5 letters of Tu0y by 1.1.69 Q&*6 o„u&yŠ S"#e35«.
v"431 90" «); T 6 is ƒ*oT T,
• š4" (
6 3/1 – #"*T^*o$ is š;6 2 and ( S56 S š56 (SD), T14 «; in ‹3901 T|T/3".
Su0y; 2 S
6
Though it is '"‹"%Ò) '" , it declines like ^*i¹. This is the license of ^"*&*4'*6 4.
• š56 1/1 – This is wo125.
• (
*‹T"3"' 7/1 ( in *us3 â'/
– from 8.2.108 T39 "yu*S *‹T"3"' «;
( 3901 7/1 2$"%-Su0> 1/2 Æ5 III/2 «

[LSK] $"%-Tu0y395 6/2 2$"%-Su0"y8"' 3/2
When there is a connection of (
and Tu0y (T, ( ‡, ( o (, q, ( 4)( with 2 and
( Su0y (S, ( ¶ (, — ð
( , ( È),(
(
2 and Su0y are the substitutes, respectively.
6
Note: this substitution has the technical name “š•”.
Since there are many possibilities in wo12, 1.3.10 3‡" ¤'4o16 25 '"4"' «( is used to
decide one wo12.
192
‹¢ *+5
Note that 1.3.10 3‡" ¤'4o16 25 '"4"' «( is used only for deciding wo12, but not for
6 should happen. Any of Æ 6 can meet with any of š 6 for
deciding the situation in which š•
6 to happen, excluding some exceptions.
š•
[LSK] %"'Ÿ21T 1 «
This example is for the combination of (
and 2. (
%"' + • 6 4.1.2 €…— '…^¶ ...( « ~ õ"T"*T^*o$"T (
( 21T 1 III/1
%"' +
%"'7• + 21T 1 8.2.66 6 7• : «
—s9
%"'5 + 21T 1 8.3.15 Ó%u "439*uy —y4/35 «
( 21T 1
%"' + 8.3.36 u" 2*% « ~ *u —y4/3Š 5
( T1
%"'2 21 6 š56 « with the help of 1.3.10 3‡"
8.4.40 Æ95 š4" ¤'4o16 25 '"4"' «(
[LSK] %"'*š49*T « Rāma performs ritual.

( + *S49*T III/1 This example is for the combination of
%"' 1/1 (
and Su0y. #*¨3" is the same.
[LSK] *íT «(
( *ST (
T+ This example is for the combination of Tu0y and Su0y.
[LSK] 2"*¹yh3 - Oh! ViJ u!, the one who has the bow called 2"¹y, be victorious.
( + —3 II/1 This example is for the combination of nasal Tu0y (4)( and Su0y.
(‹1) 2"*¹y4 8/1
•"¥' (( T +
( 2"¥')( is the example for the combination of Tu0y and 2. (
6
This š•/U •
6 section consists of five sūtras and one u"*Ty$. The first two sūtras give
6 and U•
general injunctions of š• 6 , and the last three sūtras give *4s1q, negation in special
cases.
General injunctions:
8.4.40 Æ95 6/1 š4"
6 3/1 š56 1/1 «
3/1 1/1
8.4.41 U4
6 " U56 « ~ Æ95 6/1
Negations:
( »95 6/1 Q4"' 6/1
8.4.42 4 0 ô"t"T 5/1 ( «
8.4.43 T95 6/1 *s 7/1 « ~ 4 0

( « ~ T95 6/1 4 0
8.4.44 2"T 5/1
193
6 š5«.
The next sūtra is *4s1q, prohibition to 8.4.40 Æ95 š4" 6
[*4s1q `k ']( ([*u*q `k '])( 8.4.44 2"T «( ~ T95 4 S56 *‹T"3"' (

When Tu0y is preceded by 2, ( š•
6 does not happen.
5 6
2( T6
1
S6
( « ~ T95 6/1 4 0 š56 1/1

2"T 5/1
• 2"T ( 5/1 – #"*T^*o$ is 2; Q after 2 is
( for pronunciation; in ^ku^y Í'/.
• T95 6/1 – From 8.4.43 T95 *s «; Tu0y; in v"431 90" sé/.

• 6 to happen.
4 0 – Negation for S•
• š56 1/1 – From 8.4.40 Æ95 š4"
6 š56 «.
• (
*‹T"3"' 7/1 ( in *us3 â'/
– From 8.2.108 T39 "yu*S *‹T"3"' «;
( ^%Š 6/1 Tu0yŠ 6/1 S•'

[LSK] 2"T 5/1 ( 4 0 Š"T III/1
6 1/1 ( «
6 for the Tu0y which is after 2. (
There is no S•
Note that even though there is Q* }• for this sūtra 8.4.44 2"T «( from the view of
6 because of its uS4 "'Wy, capacity of the words of sūtra, 8.4.44 2"T «( is
6 š5«,
8.4.40 Æ95 š4"
6 takes place by 8.4.40 Æ95 š4"
effective before š• 6 š5«.
6
[LSK] *u 5 « speech
*u• ( 0T… + 4™ ( 3.3.90 3—3"S3T*u•#•%“9 4™ ( « ~ Q$Ty*% $"%$1 !"3"' (
( 4
*u2 + 6
6.4.19 _95 2k¸44"* $1 S « ~ *ß™*T
(
S goes by (^þ) *4*'§"^"31 4 •*'*§$Š"A^"35 «
(When the cause is removed, the effect also goes away.)
(
4 does (
not become È because of 8.4.44 2"T «( ~ T95 4 š56
194
‹¢ *+5
[LSK] # 5 - question
( 4™ (
#• ( !/`"3"' + 3.3.90 3—3"S3T*u•#•%“9 4™ ( « ~ Q$Ty*% $"%$1 !"3"' (
( 4
#2 + 6
6.4.19 _95 2k¸44"* $1 S « ~ *ß™*T
(
4 does (
not become È because of 8.4.44 2"T «( ~ T95 4 š56
Small questions:
Give examples for each combination.
(
2+ (
( T6
2+
S6+ (
S6 + T6
( 2(
+
( S6
+
T6 + 2 (
T6 + S6
195
[*u*q `k ']( 8.4.41 6 " U56 « ~ Æ95 *‹T"3"' (

U4
(
s and »u0y are the wo12s for (
and Tu0y, respectively, when they meet.
3 6 6 3
( 6
s/» ( 6
/T ( 6
/T ( 6
s/»
1 1
( 6
s/» ( 6
s/»
U4 3/1
6 " U56
1/1
(
« ~ Æ95 6/1 *‹T"3"' 7/1
• U4 (
6 " 3/1 – #"*T^*o$ is š;6 2 and ( S56 S š56 (SD), T14 «; in ‹3901 T|T/3";
Su0y; 2 S
( thus it is !" for 5 letters of »u0y by 1.1.69 Q&*6 o„u&yŠ S"#e35«.
»6 is ƒ*oT »,
6
Though it is '"‹"%Ò) '" , it declines like ^*i¹. This is the license of ^"*&*4'*6 4.
• U56 1/1 – This is wo125.
• Æ95 6/1 – #"*T^*o$ is Æ;6 (
and Tu0y; ( T56 S Æ56 (SD), TŠ «; in v"4 1390" sé/.
S
• (
*‹T"3"' 7/1 ( in *us3 â'/
– from 8.2.108 T39 "yu*S *‹T"3"' «;
[LSK] Æ95 6/1 U4 3/1 7/1

6 " 3901 U56
1/1
(
Š"T III/1 «
When there is a connection of (
and Tu0y (T, ( ‡, ( o (, q, ( 4)( with s and
( »u0y (» (, l (, ¸ (, ñ,( &),( s (
and »u0y are the substitutes, respectively.
[LSK] %"'aé5 « Rāma is the sixth.

This example is for the combination of (
and s. (
%"' + • 6 4.1.2 €…— '…^¶ ...( « ~ õ"T"*T^*o$"T (
( sé5
%"' +
%"'7• + sé5 8.2.66 6 7• : «
—s9
%"'5 + sé5 8.3.15 Ó%u "439*uy —y4/35 «
( sé5
%"' + 8.3.36 u" 2*% « ~ *u —y4/3Š 5
(
%"'s sé5 8.4.41 U4
6 " U56 « ~ Æ95 with the help of 1.3.10 3‡" ¤'4o16 25 '"4"' «(
196
‹¢ *+5
[LSK] %"'U/$T1 « Rāma moves.

( »/$T1 This example is for the combination of
%"' + (
and »u0y.
%"'7• + »/$T1 8.2.66 6 7• : «
—s9
%"'5 + »/$T1 8.3.15 Ó%u "439*uy —y4/35«
( »/$T1 8.3.34 *u —y4/3Š 5«
%"' +
( »/$T1
%"'s + 8.4.41 U4
6 " U56 « ~ Æ95
[LSK] ^1U" « grinder

( T|S (
*^s + (
This example is for the combination of s and Tu0y.
*^s ( Sk&4y 1 to grind + T|S (agent)
( 6 S
3.1.133 Áub | …«
( T|
^1s + 6 "qyq"T$
7.3.84 "uyq"T$ 6
6 395 « ~ 0&5
( »|
^1s + 8.4.41 U4
6 " U56 « ~ Æ95
^1U| + • 6 1/1 Declined in 1/1.
^1U"
[LSK] T /$" « commentary of that

To ( + »/$" This example is for the combination of Tu0y and »u0y.
T¸ ( + »/$" 8.4.41 U4
6 " U56 « ~ Æ95
T» ( + »/$" 8.4.55 Ó*% S « ~ S% (ði"' (
[LSK] S*¨Áñ…$ 1 - Oh! ViJ u!, the one who has sudarśana cakra, may you go.
( ñ…$ 1 This example is for the combination of nasal Tu0y and »u0y.
S*¨4 +
( ñ…$ 1 8.4.41 U4
S*¨& + 6 " U56 « ~ Æ95
197
[*u*q `k ']( 8.4.42 4 ô"t" 9%4"' «( ~ Æ95 *‹T"3"' (

The U• ( ,6 but not of 4"', ( when ô"t »u0y is preceding.
6 does not happen on /T
5 6
ô »6 ( 6
/T not of 4"' (
1
( 6
s/»
( »95 5/1 Q4"' 6/1

4 0 ô"t"T 5/1 ( « Æ95 6/1 U56 1/1 *‹T"3"' 7/1
(

• 6 to happen.
• (
ô"t"T ( 5/1 – ôŠ Qt5 ô"t5 (6T) T¼"T; ( adjective to »95.
• »95 5/1 – #"*T^*o$ »6; in ^ku^y Í'/.
• Q4"' ( 6/1 – 4 4"' p*T
( Q4"' (NT);
( “4"'”( indicates “w'”,( 6/3 of ^6 -#e3
( ( the
with 4 at
6
beginning; the case ending of 6/1 is â (elided) by 7.1.39 ^"
6 %k u&"y•13"¸"."3"—"i5«.
• 6 š5«;
Æ95 6/1 – from 8.4.40 Æ95 š4" 6 in v"431 90" sé/.
• U56 1/1 – from 8.4.41 U4

6 " U56 «.
• (
*‹T"3"' 7/1 ( in *us3 â'/
– from 8.2.108 T39 "yu*S *‹T"3"' «;
( »u0"yT 5/1
[LSK] ô"t"T 5/1 ( ^%Š 6/1 Q4"'5 6/1 Æ95 6/1 U56 1/1 4 0 Š"T III/1
( «
There is no U•
6 for (
and Tu0y, other than “4"'”,( which is preceded by »u0y at the end
of ô.
[LSK] s» 1/3
( t5 1/3 « Six good people.
This is an example for the combination of ô"t » ( and . ( Because of 8.4.42 4
ô"t" 9%4"'«,( there is no U•
6 on the (
of t5.
[LSK] s» 1/3
( T1 1/3 « Those six.
This is an example for the combination of ô"t » ( and Tu0y.
198
‹¢ *+5
( done.
Now, ô$| e for the word ô"t"T is
( *$' 0( ?«
[LSK] ô"t"T 5/1 1 « He worships.
¸ ( ÆT6 … + i» (/$Ty*%/III/1
¸ ( + T1
¸ ( + »1 8.4.41 U4
6 " U56 « ~ Æ95
» ( + »1 8.4.55 Ó*% S « ~ S% (ði"' (
6
The ¸ ( of ¸ ( is »u0y, but not ô"t. By 1.4.14 *â™t ô'«,( the ô"t is ‘ of T1. Thus 8.4.42
4 ô"t" 9%4"' «( does not apply.
Another ô$| e for the word »95 is done.

[LSK] »95 5/1 *$' 0( ?« *^ yU'' «( The most genuine ghee.
( • 6 + T'^ (
*^ y + 5.3.55 Q*T2"34 1 T'*Ré4… «
#"*T^*o$ !" by 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
( + T'
*^ y + 6 q"T#6 "*T$*o$395« ~ $ (
2.4.71 ^9
( T'
*^ ys + (
8.3.101 •€"T T"o… y 5
T*}T1 « ~ 5 'kqÕ
( »'
*^ ys + 8.4.41 U4
6 " U56 « ~ Æ95
(
The s of ( ô"t, but not »u0y. Thus 8.4.42 4 ô"t" 9%4"' «( does not apply.
*^ ys is
“Q4"'”( indicates that this *4s1q is not applicable even when “4"'”,( the six case
(
singular ^6 -#e3 (
with 4 attached at the beginning, follows. The example regarding “Q4"'”(
is seen under the next u"*Ty$. u"*Ty$$"% finds a few more instances where this *4s1q is not
applicable.
199
(
This u"*Ty$ is the *4s1q of 8.4.42 4 ô"t" 9%4"' «.
(u"*Ty$')( Q4"Äu*T40%/&"*'*T u"Î' «( ~ 4 ô"t"T »95

( Æ95 U56
6 told by 8.4.42 4 ô"t" 9%4"' «( is not applicable when 4"' (6/3
This *4s1q of U• ( ^6 -(
#e3), 4u*T (number ninety), and 40%/ (city) follow »u0y. That means, U•
6 happens.
5 6
ô »6 ( 6
/T of 4"', ( 4u*T, 40%/
1
U6
(
Q4"'-4u*T-40%/&"' ( p*T 0 u"Î' 1/1
6/3
( «

• (
Q4"'-4u*T-40%/&"' (
6/3 ( 4u*T5 S 40%/ S 4"Äu*T403y5 (NT) T1s"' «;
– 4"' S ( ¬+sé/ to Æ95.
• p*T 0
• u"Î' ( 1/1 – that which had to be said.
Example for 4"' is (

( shown in the case of #"*T^*o$ ss with (
the 6th case plural ^6 -#e3.
[LSK] sÁ&"' «( of six.
( w' (
ss +
( 4"' (
ss + 7.1.55 s» (-S78š « ~ 4»6 (
ô !" for Q¹ followed by 6 etc., other than uy4"'v"4 by 1.4.17 €"*oc uy4"'v"4 1 « ~ ô' (
s¸ ( + 4"' ( 8.2.39 ði" —29›t1 «
s¸ ( + &"' ( 8.4.41 U4
6 " U56 «
( as well as this (u"þ) Q4"Äu*T40%/4"*'*T u"Î'«(
*4s1q `k for 8.4.42 4 ô"t" 9%4"' «,
excludes the case of “4"'”.(
( &"' (
s& + ( regarding 8.4.45 3%9›44"*
(u"þ) #e31 n"s"3" *4e' «, 6 6
$1 ›44"* $9 u" «
sÁ&"' (
200
‹¢ *+5
Example for 4u*T, ninety.

[LSK] sÁ&u*T5 1/1 « ninety-six.
This is a '" , whose *uœ‹ can be s¸ (-Q*q$" 4u*T «, ninety which has six extra.
6
'j'ôi9*^T°7s '" or Ò) '" can be made.
( 4u*T5
ss +
s¸ ( + 4u*T5 8.2.39 ði" —29›t1 «
s¸ ( + &u*T5 8.4.41 U4
6 " U56 «
( &u*T5
s& + 6
8.4.45 3%9›44"* 6
$1 ›44"* $9 u" «
sÁ&u*T5 or sd&u*T5
Example for 40%/, city.

[LSK] sÁ&03y5 1/3- six cities.
These can be two ZÆôs (words which are not compounded).
( 403y5
ss +
s¸ ( + 403y5 8.2.39 ði" —29›t1 «
s¸ ( + &03y5 8.4.41 U4
6 " U56 «
( &03y5
s& + 6
8.4.45 3%9›44"* 6
$1 ›44"* $9 u" «
sÁ&03y5 or sd&03y5
201
[*u*q `k ']( 8.4.43 T95 *s « ~ 4 U56 *‹T"3"' (

The U• (
6 does not happen on Tu0y when s follows.
6 7
T6 s(
1
U6
(
T95 6/1 *s 7/1 « 4 0 U56 1/1 *‹T"3"' 7/1
• T95 6/1 – #"*T^*o$ is T,6 Tu0y; in v"4 1390" sé/.
• *s 7/1 – #"*T^*o$ is s; ( in ^% â'/.
• 6 to happen.
• U56 1/1 – from 8.4.41 U4
6 " U56 «.
• (
*‹T"3"' 7/1 ( in *us3 â'/
– from 8.2.108 T39 "yu*S *‹T"3"' «;
6 '( «
[LSK] 4 0 U• 1/1
6 for Tu0y when followed by s. (

There is no U•
[LSK] eé5 - The sixth good person.

( sé5
4+ U• (
6 is negated on the 4 (Tu0y ( 8.4.43 T95 *s« ~ 4 U56
) followed by s by
Small questions: Think of each scenario.

(
s+ (
( T6
s+
»6 + (
»6 + T 6
( s(
+
( »6
+
T6 + s (
T 6 + »6
202
‹¢ *+5
‹i-( *+5 2) ô"t-—Ÿ•- *+5 “u"0/25”
(
There is another —Ÿ• by 8.4.53 ði" —2 ð*2«, which has been already seen in
6
6 ^"Š5”.
“ j
[*u*q `k ']( 8.2.39 ði" —29›t1 « ~ ôŠ

ði ( at the end of ô is replaced by —2. (
6
ô ði (
1
—2 (
( —25 1/3 Qt1 7/1 « ~ ôŠ 6/1

ði"' 6/3
• ði"' (6/3 – #e"‹"% ði;( in v"431 90" sé/; plural is because of 1.2.58 —"e"¤"3"'1$*¼4 (
RºuS4'ÕT%Š"' «(
• —25 1/3 – #e"‹"% — ; ( this is wo12.
• Qt1 7/1 – in Q*q$%&1 â'/.
• ôŠ 6/1 – from Q*q$"% `k 8.1.16 ôŠ «; in ¬q1 sé/ to Qt1 .
( —25 1/3 Š56 III/3 «

[LSK] ô"t1 7/1 ði"' 6/3
At the end of ô, ði ( is replaced by —2. (
[LSK] u"0/25 -
u"S5 25 u"0/25 (6T)« The lord of speech.
( ™ +
u"S + ( 2 + •6 6
2.2.8 sé/ « ~ '" 5 T°7s5 6 ‹
^"
#"*T^*o$ !" by 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
( 2
u"S + 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
Even after the elision of ^6 , ( by 1.4.14 *â™t
6 (
ô'«,( the S has ô"t•.
u"$ ( + 2 8.2.30 S95 $65 « ~ ôŠ Qt1
( 2
u"0 + 8.2.39 ði" —29›t1 « ~ ôŠ with 1.1.50 v"4›1 t%T'5 «
203
‹i-( *+5 3) Q44"*

6 $- *+5 “‘T %"*%5”
6
[*u*q `k ']( 8.4.45 6 $1 ›44"*

3%9›44"* 6 $9 u" « ~ ô"tŠ *‹T"3"' (
3% (at the end of ô is replaced by the nasal letter closest to it when a nasal letter
follows.
6 7
ô 3% ( 6
Q44"* $
1
Optional
6
Q44"* $
3%5 6/1 Q44"*

6 $1 7/1 Q44"*
6 (
$5 1/1 u" 0 « ~ ôŠ 6/1 *‹T"3"' 7/1
• 3%5 6/1 – #e"‹"% 3%;( in v"431 90" sé/.
• 6
Q44"* 6
$1 7/1 – Q44"* $ is !" defined by 1.1.8 'Ó 6
6 4"* $"uS49›44"* $5«; in ^% â'/.
• 6
Q44"* $5 1/1 – This is wo12.
• u" 0 – This rule is optional.
• ô"tŠ 6/1 – from 8.4.42 4 ô"t" 9%4"'«;( this is adjective to 3%5 .
[LSK] 3%5 6/1 ô"tŠ 6/1 Q44"*

6 $1 7/1 ^%1 7/1 Q44"*
6 (
$5 1/1 u" 0 Š"T III/1 «
Nasal letter is the substitute in the place of 3% (at the end of ô when Q44"*
6 $
follows.
6
Among the many letters which are termed Q44"* $, the substitute is decided by
1.1.50 v"4›1 t%T'5«.
For 3, ( u, ( and i,( the Q44"*
6 $ letters are 3,• ( u,• ( and i• ,( respectively. As for % (and
sibilants (2, ( s, ( and ),( there is no Q44"*
6 $ letter. For the letters in the five u0ys, the
(
substitute should be the 5th (nasal) letter of the u0y to which the v"*44 belongs.
204
‹¢ *+5
[LSK] ‘T %6 "*%5 « This ViJ u (This can be either 'Æô or ZÆô.)

‘TŠ '%6 "*%5 ‘T %6 "*%5 (6T)« or ‘s5 '%6 "*%5 ‘T %6 "*%5 (KT)«
( '%6 "*% + • 6 2.2.8 sé/ « ~ '" 5 T°7s5
‘To ( + ™ + 6 6 ‹
^"
#"*T^*o$ !" by 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
‘To ( + '%6 "*% 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
Even after the elision of ^6 , ( by 1.4.14 *â™t
6 (
ô'«,( the S has ô"t•.
‘To ( + '%6 "*% 8.2.39 ði" —29›t1 « ~ ôŠ with 1.1.50 v"4›1 t%T'5 «
( '%6 "*%
‘T4 + 6
8.4.45 3%9›44"* 6
$1 ›44"* $9 u" « ~ ô"tŠ
‘T %6 "*%
6
^“1 ‘T/%"*%
205
6
This u"*Ty$ gives *4e for Q44"* $ wo12, which is optional in the sūtra.
(u"*Ty$')( #e31 n"s"3" *4e' «( ~ 3%5 Q44"*

6 6
$1 Q44"* $5 *‹T"3"' (
In non-Vedic literature, nasal letter is the substitute compulsorily when nasal-
beginning suffix follows.
6 7
ô 3% ( 6
Q44"* $ #e3
1
Compulsory (*4e)
6
Q44"* $
( *4e' 0( « ~ ôŠ 6/1 3%5 6/1 Q44"*

#e31 7/1 n"s"3"' 7/1 6 $1 7/1 Q44"*
6 $5 1/1
• #e31 7/1 – suffix; in ^% â'/.
• n"s"3"' (7/1 – #"*T^*o$ is n"s", indicating non-Vedic literature; in *us3 â'/.
• *4e' 0( – *¨3"*u21s& (adverb); always; contrasting to “u"” in 8.4.45 3%9›44"*
6 6
$1 ›44"* $9 u"«.
[LSK] T "`' «(
(
TT #'"&' (
QŠ p*T T "`' «( That much. That for which the measurement is that.
To ( + • 6 + '"`S ( (
5.2.37 #'"&1 Ò3 —o1Ø'"`S5 « ~ To ( QŠ
#"*T^*o$ !" by 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
To ( + '"` 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
Even after $,( because of the ^6 -#e3,
( 6
there is ô !" for To ( by 1.4.1.4 *â™t ô' «(
To ( + '"` 8.2.39 ði" —29›t1« ~ ôŠ with 1.1.50 v"4›1 t%T'5 «
( '"`
T4 + 6
u•$*®$ (optional) Q44"* 6
$ was #"â by 8.4.45 3%9›44"* 6
$1 ›44"* $9 u" «,
(
but it is made compulsory by (u"þ) #e31 n"s"3" *4e' «.
6
By 1.1.50 v"4›1 t%T'5«, Q44"* $ of o• o ( is determined as o• 4. (
T "`
206
‹¢ *+5
(
[LSK] *S 3' -
*So ( ‘u p*T *S 3' «( That which is only consciousness.
( '3» (
*ST + 4.3.144 *4e u|}2%"*o85 « ~ '3»19(
*So ( + '3 8.2.39 ði" —29›t1« ~ ôŠ with 1.1.50 v"4›1 t%T'5 «
( '3
*S4 + 6
u•$*®$ (optional) Q44"* 6
$ was #"â by 8.4.45 3%9›44"* 6
$1 ›44"* $9 u" «,
(
but it is made compulsory by (u"þ) #e31 n"s"3" *4e' «.
6
By 1.1.50 v"4›1 t%T'5«, Q44"* $ of o• o ( is determined as o• 4. (
*S 3
19
Here, by 390*un"0 (dividing a sūtra into two in order to get pU* *}), only “*4e'”( is taken to get
'3» ( for this #"*T^*o$ in the sense of €"‡Ë.
207
‹i-( *+5 4) Last section
[*u*q `k ']( 8.4.60 T9*iy « ~ ^% u&y5 *‹T"3"' (

Tu0y is replaced by ^% u&y when i ( follows.
6 7
Tu0y i(
1
^% u&y
(
T95 6/1 *i 7/1 « ~ ^% u&y5 1/1 *‹T"3"' 7/1
• T95 6/1 – #"*T^*o$ is T,6 ƒ*oT T( , ( which is !" for Tu0y by 1.1.69 Q&*6 o„u&yŠ S"#e35«;
in v"431 90" sé/.
• *i 7/1 – #"*T^*o$ is i;( in ^% â'/.
• ^% u&y5 1/1 – ^%Š u&y5 ^% u&y5 (6T)« u&y of what follows; this is wo12.
Since what follows is only i,( the ^% u&y can only be i ( or i• .(
[LSK] Tu0yŠ 6/1 i$"%1 7/1 ^%1 7/1 ^% u&y5 1/1 «

u&y of what follows is the substitute in the place of Tu0y when i ( follows.
To decide the wo12, which is either i ( or i• ,( 1.1.50 v"4›1 t%T'5« is applied. Since
( non-nasal (T, ( ‡, ( o (, q),( the wo12 is also non-nasal
4"* $", nose, is also v"4, when the v"*44 is
( nasal (4),( the wo12 is also nasal (i• )( .
(i),( while when the v"*44 is
208
‹¢ *+5
This is the example for non-nasal v"*44. (

[LSK] T 35 1/1 «
TŠ i35 T 35 (6T)« Resolution of that.
( i3 + • 6
To ( + ™ + They are compounded.
To ( + i3 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
6
To ( has the status of ô by by 1.4.14 *â™t ô' «(
To ( + i3 8.2.39 ði" —29›t1« ~ ôŠ with 1.1.50 v"4›1 t%T'5 «
Ti ( + i3 8.4.60 T9*iy « ~ ^% u&y5
By 1.1.50 v"4›1 t%T'5«, u&y to o ( is non-nasal i.(
This is the example for nasal v"*44. (

[LSK] *uÒ"i• ( *iÓ*T « The learned person writes.
( + *iÓ*T III/1
*uÒ"4 6/1
*uÒ"i• ( + *iÓ*T 8.4.60 T9*iy « ~ ^% u&y5
( nasal i• .(
By 1.1.50 v"4›1 t%T'5«, u&y to 4 is
[LSK] 4Š 6/1 Q44"*
6 $5 1/1 i5 1/1 -
For 4$"%, wo12 should be nasal i$"%.
Because of the limitation of the font-set of printing, what is usually seen is
6
“*uÒ"* • Ó*T”, which looks as though the p of *iÓ*T is Q44"* $. The correct understanding is
( *uÒ"4 becomes
that the 4 of ( i• .(
209
[*u*q `k ']( y « ~ u&y5 *‹T"3"' (

8.4.61 ƒo5 v"Æf95 ^kuŠ
^ku y u&y (‡)( is the substitute for the first letter of v" and Æf when
( they are preceded
by ƒo (.
5 6
ƒo ( v"/Æf (
1
^ku y u&y
ƒo5 5/1 v"Æf95 6/2 ^kuŠ (

y 6/1 « u&y5 1/1 *‹T"3"' 7/1
• ƒo5 5/1 – ƒo ( is a particle found in #"*o0&, used as ƒ^ 0y with v" and Æf; ( in ^ku^y Í'/.
• ( v"Æf… (ID) T395«; two q"Ts6 , i.e., v" and Æf; ( in v"431 90"
v"Æf95 6/2 – v" S Æf S
sé/.
• y 6/1 – in ¬+sé/ to u&y5.
^kuŠ
• y
u&y5 1/1 – This is wo12. Together, ^kuŠ u&y5 indicates the u&y of preceding letter.
[LSK] ƒo5 5/1 ^%395 6/2 v"Æf95 6/2 ^ku y u&y5 1/1 -
(
u&y of what is preceding is the substitute for v" and Æf when ƒo ( precedes.
(
Examples to be seen here are ƒo ( + v"4' (v" (
+ Þ»6 () and ƒo ( + Æf4' (Æf ( Þ»6 ().
+
The next two ^*%n"s" `k s are required for applying this sūtra.
210
‹¢ *+5
y «.
This sūtra has already been studied in this book with 1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
The 5th case is understood as *oÏ3901 ^Í'/. However, the doubt is which direction it
should be. The next ^*%n"s" `k answers this doubt regarding ^Í'/-ending word.
[^*%n"s" `k ']( 6
1.1.67 T¼"*oe§%Š « ~ *4*oýU"T (
( p*T 0 ƒ§%Š 6/1 « ~ *4*oýU"T 5/1

T¼"T 5/1 (

• (
• (
gap; after *un*´*u^*%&"', in *oÏ3901 ^Í'/, '"4"*q$%& to T¼"T. (
• ƒ§%Š 6/1 – “$"3y'”( is supplied; in ¬+sé/.
( $"3y' 1/1
[LSK] ^Í'/*4oË241 3/1 *uq/3'"4' 1/1 ( u&"yt%1& 3/1 QZu*‹TŠ 6/1 ^%Š 6/1 (=ƒ§%Š 6/1) !13' (
1/1
«
An effect, which is being enjoined by presented a 5th case-ending word, should be
( understood to be v" and Æf by

In the case of ƒo ( + v", and ƒo ( + Æf, ( the v"*44 is (
6
1.1.67 T¼"*oe§%Š «.
5 6
ƒo ( v"/Æf (
1
^ku y u&y
(
Now, since the v"*44 has more than one letter, we need ^*%n"s" `k to decide which
letter is to be substituted among all the letters in v" and Æf. ( The next sūtra gives the
solution.
211
The next ^*%n"s" `k is from the section which teaches where the replacement should
happen. This is Qû"o to 1.1.52 Qi9›•Š «.
[^*%n"s" `k ']( 1.1.54 wo15 ^%Š « ~ Qi5

When substitute is enjoined with a word in 5th case, the first letter after that should
be the place of replacement.
1
wo12
wo15 6/1 ^%Š 6/1 « ~ Qi5 6/1

• wo15 6/1 – #"*T^*o$ is w*o; adjective to Qi5.
• ^%Š 6/1 – in ¬+sé/ to wo15.
• Qi5 6/1 – #"*T^*o$ is Qi,( any letter ; in v"4 1390" sé/.
( *u*‹T' 1/1
[LSK] ^%Š 6/1 3T 1/1 ( TT 1/1
( TŠ 6/1 wo15 6/1 R9j' 1/1
( «
That which is enjoined for what follows (after a 5th case-ending word) is to be
understood (as enjoined) for the beginning of that.
6
“^%Š” is a synonym to “ƒ§%Š” which comes from 1.1.67 T¼"*oe§%Š «. When a
substitute is enjoined with a word in 5th case, the place of operation is known as “ƒ§%Š
v4”1 .
[LSK] p*T 0 Š 6/1 ‡5 1/1 -

Coming back to the example of ƒo ( + v"4', ( since ƒo ( is told in the 5th case, , ( the
( 1.1.54 wo15 ^%Š «. Detail is given in the following.
( the v"*44 by
beginning letter of v"4' is
ƒo ( + v"4' (
212
‹¢ *+5
(
ƒo ( + ‡ ‡"4' ( 8.4.61 ƒo5 v"Æf95 ^kuŠ
y «
6
1.1.67 T¼"*oe§%Š «
1.1.53 wo15 ^%Š «
1.1.50 v"4›1 t%T'5«
^ku y u&y, similar letter to o ( is the wo12. u&ys of o ( are T, ( ‡, ( o (, q, ( 4. ( In order to decide one
wo12, 1.1.50 v"4›1 t%T'5« is used. Since v"4 of ( ot"5, which are the same as all the u&ys
is
6
of o (, 0&T5 Qt%T'5, most similar by the #3É has to be considered. ( hard (Q•9s) and
is
aspirated ('‹"#"&). Among the u&ys of o ( (T, ( ‡, ( o (, q, ( 4),( that which is Q•9s and '‹"#"& is ‡. (
Thus ( replaced by ‡. (
is
These following two sūtras are studied now.
ƒo ( + ‡"4' ( 8.4.65 ð%9 ð*% u&Ë « ~ u" i9^5
( ‡"4' (
ƒT + 8.4.55 Ó*% S « ~ ði" S% (
213
[*u*q `k ']( (
8.4.65 ð%9 ð*% u&Ë« ~ ‹i5 QÕT%Š"' i9^5 *‹T"3"' (
ð% (is optionally elided when preceded by ‹i ( and followed by u&y ð%.(
5 6 7
‹i ( ð% ( u&y ð% (
1
optional
i9^
ð%5 6/1 ð*% 7/1 u&Ë 7/1 « ~ ‹i5 5/1 i9^5 1/1 QÕT%Š"' 0( *‹T"3"' 7/1
(

• ð%5 6/1 – #e"‹"%5 ð%;( letters in 1st to 4th of the classes, and sibilants; in v"431 90" sé/.
• ð*% 7/1 – #e"‹"%5 ð%;( letters in 1st to 4th of the classes, and sibilants; in ^% â'/.
• u&Ë 7/1 – This is adjective to ð*%. (As they are identical, To"*o*u*q is not required.)
• ‹i5 5/1 – From 8.4.64 ‹i9 3'" 3*' i9^5 «; #e"‹"%5 ‹i;( in ^ku^y Í'/.
• i9^5 1/1 – From 8.4.64 ‹i9 3'" 3*' i9^5 «; this is wo12.
• QÕT%Š"' (0 – From 8.4.62 ð39 ‹9›ÕT%Š"' «;
( synonym to “u"”.
Since sūtra literature has to have Q®"“%•, minimum usage of letters, one may think
that "QÕT%Š"'"( takes too many syllables compared to its one-syllabled synonym, "u"".
However, it is said “^3"y32²"4" i"•u0…%uSS"y 4"*L3T1, the discussion for smallness or bigness
among synonyms is not respected.” For this reason, we do not question when we see
"QÕT%Š"'",( "*un"s"", etc., for "u"", or "#n|*T" for "w*o".
[LSK] ‹i5 5/1 ^%Š 6/1 ð%5 6/1 u" 0 i9^5 1/1 u&Ë 7/1 ð*% 7/1 -
i9^ is optionally the substitute in the place of ð% (which is after ‹i,( when followed
by ð%,( which is u&y to the preceding ð% (.
ƒo ( + v"4' (
(
ƒo ( + ‡ ‡"4' ( 8.4.61 ƒo5 v"Æf95 ^kuŠ
y «
ƒo ( + ‡"4' ( 8.4.65 ð%9 ð*% u&Ë « ~ ‹i5 i9^5 QÕT%Š"' (
214
‹¢ *+5
[*u*q `k ']( 8.4.55 Ó*% S « ~ ði" S% ( *‹T"3"' (

ði ( is replaced by S% ((1st of the class and sibilants), when it is followed by Ó% ((hard
consonant).
6 7
ði ( Ó% (
1
S% (
( S% 1/1
Ó*% 7/1 S 0 « ~ ði"' 6/3 ( (
*‹T"3"' 7/1
• Ó*% 7/1 – #e"‹"%5 Ó%;( all the hard consonants; in ^% â'/.
• S 0 – Connecting to the previous sūtra.
• (
ði"' 6/3 (
– From 8.4.53 ði" —2 ð*2 «; #e"‹"%5 ði;( in v"431 90" sé/; plural is because of
(
1.2.58 —"e"¤"3"'1$*¼4 RºuS4'ÕT%Š"' «(
• S% (1/1 – 8.4.54 Q8" 1 SSy «; #e"‹"%5 S%;( letters in 1st of the class; this is wo12.
( S%5 1/3 Š56 III/3 «

[LSK] Ó*% 7/1 ði"' 6/3
S% (are the substitutes in the place of ði ( when they are followed by Ó%.(
[LSK] p*T 0 ƒo5 5/1 oŠ 6/1 T5 1/1 «
( the substitute in the place of o ( of ƒo (.
Thus T is
( «
[LSK] ƒô"4' 1/1
ƒo ( + v"4' (
(
ƒo ( + ‡ ‡"4' ( 8.4.61 ƒo5 v"Æf95 ^kuŠ
y «
ƒo ( + ‡"4' ( 8.4.65 ð%9 ð*% u&Ë « ~ ‹i5 i9^5 QÕT%Š"' (
( ‡"4' (
ƒT + ( (
8.4.55 Ó*% S « ~ ði"' S%
ƒô"4' (
( Q* }. Instead,
Note that in the view of 8.4.55 Ó*% S «, ‡ is ( the ^%*4*'§.
is
215
i9^-Qn"u-^“1 (When the optional i9^ by 8.4.65 ð%9 ð*% u&Ë « is not taken,)
ƒo ( + v"4' (
(
ƒo ( + ‡ ‡"4' ( 8.4.61 ƒo5 v"Æf95 ^kuŠ
y «
( g‡"4' (
ƒT + ( (
8.4.55 Ó*% S « ~ ði"' S%
ƒ<g‡"4' (
(
Note that 8.4.55 Ó*% S « does not work for the first ‡ because it is a $"3y of 8.4.61 ƒo5
v"Æf95 ^kuŠ (
y « and it is Q* }uT from the view of 8.4.55 Ó*% S «.
( -
[LSK] ƒ§f4'1/1
ƒo ( + Æf4' (
(
ƒo ( + ‡ Tf4' ( 8.4.61 ƒo5 v"Æf95 ^kuŠ
y «
ƒo ( + Tf4' ( 8.4.65 ð%9 ð*% u&Ë « ~ ‹i5 i9^5 QÕT%Š"' (
( Tf4' ( 8.4.55 Ó*% S « ~ ði"' S%
ƒT + ( (
ƒ§f4' (
^“1 (When the optional i9^ by 8.4.65 ð%9 ð*% u&Ë « is not taken,)
ƒo ( + Æf4' (
(
ƒo ( + ‡ Tf4' ( 8.4.61 ƒo5 v"Æf95 ^kuŠ
y «
( gTf4' ( 8.4.55 Ó*% S « ~ ði"' S%
ƒT + ( (
ƒ<gTf4' (
(
Again, 8.4.55 Ó*% S « does not work for the ‡ because it is a $"3y of 8.4.61 ƒo5 v"Æf95
^kuŠ (
y « and it is Q* }uT from the view of 8.4.55 Ó*% S «.
216
‹¢ *+5
[*u*q `k ']( 8.4.62 ð39 ‹9›ÕT%Š"' «( ~ ^kuŠ

y u&y5 *‹T"3"' (
When preceded by ð3, ( ‹ ( is optionally replaced by the 4th of the class of what is
preceding.
5 6
ð3 ( ‹(
1
Optional
^ku y u&y
ð35 5/1 ‹5 6/1 QÕT%Š"' 0( «~ ^kuŠ (

y 6/1 u&y5 1/1 *‹T"3"' 7/1
• ð35 5/1 – #e"‹"%5 ð3; ( all the letters of 1st to 4th of the class; in ^ku^y Í'/.
• ‹5 6/1 – #"*T^*o$ is ‹ (; in v"431 90" sé/.
• QÕT%Š"' 0( – This *u*q is optional.
• y 6/1 – Of the previous, that which precedes; in ¬+1 sé/ to u&y5.
^kuŠ
• u&y5 1/1 – As defined by 1.1.9 TÞ ( this is wo12.
6 "Š#3É u&y' «;
( ‹Š 6/1 u" 0 ^ku y u&y5 1/1 «

[LSK] ð35 5/1 ^%Š 6/1
^ku y u&y is optionally the substitute in the place of ‹ ( which is after ð3. (
[LSK] 4"oŠ 6/1 •9sŠ 6/1 u"%Š 6/1 '‹"#"&Š 6/1 T"”25 1/1 u0yST‡6 5y 1/1 «
What is similar to (‹$"%, which is) 4"o, •9s, u"% (together it means “soft consonant”)
and '‹"#"& is the 4th of the class.
(
^*%n"s" `k 1.1.50 v"4›1 t%T'5« is used to decide the wo12. Since ð3 covers all the five
v"4, the u&y of ^ku y should be the five letters of each of the five u0ys. Because the choices for
wo12 are among u&ys, the letters which have the same v"4 and w8t%#3É, the wo12
should be decided by the R"ã#3É. Since R"ã#3É of ‹ ( is soft (4"o, •9s, and u"%) and aspirated
('‹"#"&), the similar letter with the R"ã#3É among u&ys is the 4th of the class.
217
[LSK] u"Ï•*%5 1/1 « u"Ï‹*%5 1/1 -

u"*S ‹*%5 ^*Á¸T5 u"Ï•*%5 (6T) « The one who is skilled in speech.
( *™ + ‹*% + • 6
u"S +
(
u"S ‹*%
u"$ ( ‹*% 8.2.30 S95 $65 « ~ ð*i ôŠ Qt1 S
(
u"0 ‹*% 8.2.39 ði" —29›t1 « ~ ôŠ
(
u"0 •*% 8.4.62 ð39 ‹9›ÕT%Š"' «( with 1.1.50 v"4›1 t%T'5 «
u"Ï•*%
^ku y u&y-Qn"u-^“1, in the altanative case of absence of ^ku y u&y,

u"Ï‹*%
Examples for 8.4.62 ð39 ‹9›ÕT%Š"' «(

In Su0y ù$"i9› €o/•‚y 6T5 «
In »u0y Š"¸ (O€5
In Tu0y %"o (qkT 1
In û0y QhÆ/
218
‹¢ *+5
[*u*q `k ']( 8.4.63 2Ÿ¶9›*» « ~ ð35 QÕT%Š"' ( *‹T"3"' (

When preceded by ð3, ( and followed by Q» (, 2 is
( optionally replaced by ¶ (.
5 6 7
ð3 ( 2( Q» (
1
Optional
¶(
25 6/1 ¶5 1/1 Q*» 7/1 « ~ ð35 5/1 QÕT%Š"' 0( *‹T"3"' 7/1

(

• 25 6/1 – #"*T^*o$ is 2; ( in v"431 90" sé/;
• ¶5 1/1 – This is wo12; Q is for pronunciation.
• Q*» 7/1 – #e"‹"%5 Q» (; in ^% â'/.
• ð35 5/1 – #e"‹"%5 ð3; ( all the letters of 1st to 4th of the class; in ^ku^y Í'/.
• QÕT%Š"' 0( – This *u*q is optional.
[LSK] ð35 5/1 ^%Š 6/1 2Š 6/1 ¶5 1/1 u" 0 Q*» 7/1 «
(
¶ ( is optionally the substitute in the place of 2 which is preceded by ð3, ( and
followed by Q» (.
[LSK] “To ( *2u5” p*T 0 Q` 0 oŠ 6/1 š•1
6 4 3/1 —$"%1 7/1 $| T1 7/1 “Ó*% S” p*T 0 —$"%Š 6/1 S$"%5 1/1 «
6 by 8.4.40 Æ95 š4"
In this example of “To ( *2u5”, by the š• 6 o ( becomes —, ( and by
6 š5«,
(
8.4.55 Ó*% S«, — becomes S. (
[LSK]T*•u5, T*ˆ2u5 «
To ( *2u5
(
T— *2u5 6 š5«
8.4.40 Æ95 š4" 6
(
TS *2u5 8.4.55 Ó*% S«
(
TS *¶u5 8.4.63 2Ÿ¶9›*» « ~ ð35 QÕT%Š"' (
T*•u5
¶•"n"u^“1 T*ˆ2u5
219
u"*Ty$$"% extends the scope of the previous sūtra.
(u"*Ty$')( ¶•' Q'/*T

( u"Î' «(
With reference to 8.4.63 2Ÿ¶9›*»«, instead of Q*», Q*' should have been told.
5 6 7
ð3 ( 2( Q' (
1
Optional
¶(
( Q*' 7/1 p*T 0 u"Î' 1/1

¶•' 1/1 ( «

• (
¶•' 1/1 – ¶ (-wo12 in 8.4.63 2Ÿ¶9›*» «.
• Q*' 7/1 – #e"‹"%5 Q'; ( compare to Q» (, i ( and È, ( ', ( ™,( &, ( 4 are
( extra; in ^% â'/.
• p*T 0 – this puts “Q*'” in the quote.
• (
u"Î' 1/1 – “should be told”; '"4"*q$%& and subjective complement to ¶•'. (
[LSK] Ti9$1 4 3/1-

To ( +9$1 4
(
T— +9$1 4 6 š5«
8.4.40 Æ95 š4" 6
(
TS +9$1 4 8.4.55 Ó*% S«
(
TS ji9$1 4 (u"þ) ¶•''/*T u"Î' «(
( followed by i,( which is not
By 8.4.63 2Ÿ¶9›*»«, this case is not covered because 2 is
in Q» (, but in Q'. (
Notes:
There is a !"^$ (indication) of this u"*Ty$ in Pā ini’s sūtra itself. The sūtra 5.1.134
( Ts1 6 « requires Q*' as ^%*4*'§.
09`S%4"•i"•"e"$"%Tou1
(
To avoid ¶• within a word such as “*u%k' (abundant)”, “ô"t"T”( from 8.4.59 u"
ô"tŠ can be taken and changed to 5th case ending by *un*´*u^*%&"', to qualify ð35.
220
‹¢ *+5
‹i-( *+5 5) Q4€"%-

6 *+5 “‹*% u¡1”
[*u*q `k ']( 6
8.3.23 '9›4€"%5 « ~ ôŠ ‹*i *‹T"3"' (
( replaced by Q4€"%
ô"t ' is 6 when followed by a consonant.
6 7
ô '( ‹i (
1
6
Q4€"%
'5 6/1 Q4€"%5

6 1/1
(
« ~ ôŠ 6/1 ‹*i 7/1 *‹T"3"' 7/1
• '5 6/1 – #"*T^*o$ is '; ( adjective to ôŠ;
• 6
Q4€"%5 1/1 – This is wo12.
• ôŠ 6/1 – from 8.2.108; in v"431 90" sé/; with its *u21s& (adjective) '5 6/1, Tot*u*q by
(
1.1.72 314*u*qÆotŠ« is applied, and they together mean “'-QtŠ ôŠ”.
• ‹*i 7/1 – #e"‹"%5 ‹i;( in ^% â'/.
[LSK] '"tŠ 6/1 ôŠ 6/1 Q4€"%5

6 1/1
‹*i 7/1 «
6
Q4€"% (
is the substitute in the place of the last letter of '-ending ô, when ‹i (
follows.
[LSK] ‹*% 2/1 u¡1 I/1 - I salute Hari.
(
‹*%' u¡1 (
‹*%' gets 6
ô !" by 1.4.14 *â™t ( is thereby '-ending
ô' «, ( ô.
‹*% u¡1 6
8.3.23 '9›4€"%5 « with 1.1.52 Qi9›•Š «
6
Note: Q4€"% is a letter, whose v"4 is 4"* $" only, according to *2“" `k . (ref. u|*§ of 1.1.9
( Q4€"%
6 Š"#3É u&y' «)
TÞ 6 is one of the letters called Q390u"‹s, which are not enumerated in
6
'"‹1Â% `k because they do not appear independently. Q390u"‹s are: Q4€"%5, *u 0y5, *—;"'ki/35,
ƒ^:"4/35, and 3' (special sound used in Vedic chanting). Even though they are not
enumerated in '"‹1Â% `k , they are included in #e"‹"% Q» ( and 2i.( [Bh 1-138]
221
6
The next sūtra also gives Q4€"%.
[*u*q `k ']( 8.3.24 4š"ô"tŠ ð*i « ~ '5 Q4€"%5

6 *‹T"3"' (
(
Qô"t ' and (
4 are 6
replaced by Q4€"% when followed by ði.(
6 7
ô ( (
'/4 ði (
1
6
Q4€"%
45 6/1 S 0 Qô"tŠ 6/1 ð*i 7/1 « ~ '5 6/1 Q4€"%5

6 1/1
(
*‹T"3"' 7/1
• 45 6/1 – #"*T^*o$ is 4; ( in v"431 90" sé/.
• S 0 – connecting with '5; 45 S '5 S.
• Qô"tŠ 6/1 – 4 ô"t5 Qô"t5 (NT), TŠ; adjective to 45 and '5.
• ð*i 7/1 – #e"‹"%5 ði;( in ^% â'/.
• '5 6/1 – #"*T^*o$ is '; ( in v"431 90" sé/.
• 6
Q4€"%5 1/1 – This is wo12.
[LSK] 4Š 6/1 'Š 6/1 S 0 Qô"tŠ 6/1 ð*i 7/1 Q4€"%5

6 1/1
«
6
Q4€"% (
is the substitute in the place of 4 and (
' which is Qô"t, when ði ( follows.
Small question:
What are the differences between this sūtra and the previous sūtra?
222
‹¢ *+5
By the nature of the sūtra, two types of example are to be presented. The example
( presented first.
for 4 is
1/3
[LSK] 32"* «
( fame) in 1/3.
#"*T^*o$ 32 (n.
( — 1/3
32 + ( 4.1.2 €…— …
( « ~ #"*T^*o$"T (
( *2
32 + 7.1.20 —Ÿ2 95 *2 « ~ 4^ 6 $"T (
The wo12 *2 gains uy4"'v"4 !" by 1.1.42 *2 uy4"'v"4' «(
•6 ( +
32 4' ( p 7.1.72 4^ 6 $Š ðiS5 « ~ 4'6 (
32" 4 ( p( 6.4.10 "t'‹T5 390Š « ~ ƒ^q"3"5 o/•5y

32"* 6
8.3.24 4š"ô"tŠ ð*i « ~ '5 Q4€"%5
This is an example for Qô"t '$"%:

[LSK] w¨ŠT1 III/1 «
( i|» (/$Ty*%/III/1
w™ ( + ¨' +
( Š + T1
w + ¨' + 3.1.33 ŠT" / i| »95 «
w¨ŠT1 6
8.3.24 4š"ô"tŠ ð*i « ~ '5 Q4€"%5
ô$| e is done.
[LSK] ð*i p*T *$'?( 'ÕT1 -
(
Why is ð*i told? For 'ÕT1 = '4 (4A) to consider+ i» (/$Ty*%/III/1.
( followed by 3, ( which is not ði.( Thus Q4€"%
Qô"t 4 is 6 does not happen.
223
6
The next sūtra produces a change for Q4€"%.
[*u*q `k ']( 6
8.4.58 Q4€"%Š 3*3 ^% u&y5 « ~ *‹T"3"' (
6
Q4€"% is replaced by ^% u&y when followed by 33. (
6 7
6
Q4€"% 33 (
1
^% u&y
6
Q4€"%Š 6/1
(
3*3 7/1 ^% u&y5 1/1 « *‹T"3"' 7/1
• 6
Q4€"%Š 6/1 – in v"431 90" sé/.
• 3*3 7/1 – #e"‹"%5 33; ( all the consonants except for sibilants; in ^% â'/.
• ^% u&y5 1/1 – ^%Š u&y5 ^% u&y5 (6T); this is wo12.
6
wo12 is decided by 1.1.50 v"4›1 t%T'5«. Because the v"4 of Q4€"% is 4"* $", letters
which are the 5th of each class will be the wo12 for u"0*3-u&ys (25 letters which belong to five
classes). As for 3, ( u, ( and i,( the Q44"*
6 $ 3,• ( u,• ( and i• ( will be the wo12, respectively. %1Ý does
6
not have u&y, thus Q4€"% remains as it is.
[LSK] ÐU' 0( «
Clear. No explanation is required. No Q4u|6 *§ is required to understand this sūtra.
[LSK] 2"t5 -
( ´
2' + 3.2.102 *4é" « 7.2.27 u" o"t2"t^k&oy ÆÐU•ü!â"5 «
( T
2"' + 6
6.4.15 Q44"* $Š *ò-ði95 *ß™*T «
2"T 6
8.3.24 4š"ô"tŠ ð*i « ~ Q4€"%5
2"t 6
8.4.58 Q4€"%Š 3*3 ^% u&y5 « with ^*%n"s" 1.1.50 v"4›1 t%T'5«
224
‹¢ *+5
[*u*q `k ']( 8.4.59 u" ô"tŠ « ~ Q4€"%Š

6 3*3 ^% u&y5 *‹T"3"' (
6
The replacement of Q4€"% is optional at the end of ô.
6 7
ô 6
Q4€"% 33 (
1
Optional
^% u&y
u" 0 ô"tŠ 6/1 « ~ Q4€"%Š

6 6/1
(
3*3 7/1 ^% u&y5 1/1 *‹T"3"' 7/1
• u" 0 – This indicates that this rule is optional.
• 6
ô"tŠ 6/1 – ôŠ Qt5 ô"t5 (6T), TŠ «; adjective to Q4€"%Š .
• 6
Q4€"%Š 6/1 – in v"431 90" sé/.
• 3*3 7/1 – #e"‹"%5 33; ( all the consonants except for sibilants; in ^% â'/.
• ^% u&y5 1/1 – ^%Š u&y5 ^% u&y5 (6T); this is wo12.
[LSK] • %9*s, • $%9*s 1/1, II/1 -

(
•' $%9*s ( a Rtô,
•' is 6 followed by $,( which is 33. (
• $%9*s 6
8.3.23 '9›4€"%5 « ~ ôŠ ‹*i
•™ ( $%9*s 6
8.4.59 u" ô"tŠ « ~ Q4€"%Š 3*3 ^% u&y5, with ^*%n"s" 1.1.50 v"4›1 t%T'5«
^“1
• $%9*s
225
6
The next sūtra is Qû"o to 8.3.23 '9›4€"%5« 6
and negates Q4€"%, in order to make the
word “ l"—”.(
[*u*q `k ']( 8.3.25 '9 %"*— '5 ò… « ~ '5 *‹T"3"' (

(
The ' of ( replaced by ', ( not by Q4€"%,
' is 6 (
when followed by *ò^-ending ( .6
%"— q"T
6 7
'( ( *ò^ (
%"— +
1
'(
(
'5 1/1 %"*— 7/1 '5 6/1 ò… 7/1 « ~ '5 6/1 *‹T"3"' 7/1
• ( the wo12, and the Q after ' is
'5 1/1 – This is wo12; ' is ( for pronunciation, ƒí"%&"‡5y .
• %"*— 7/1 – #"*T^*o$ is %"—20( ; q"T 6 %"— o/â…

( (1A) to shine; in ^% â'/.
• '5 6/1 – #"*T^*o$ is '; ( indicating ' ƒ^
( 0y before %"—; ( in ¬+1 sé/ to '5.
• (
ò… 7/1 #"*T^*o$ is *ò; indicating *ò^ #e3; in ^% â'/.
• '5 6/1 – #"*T^*o$ is '; ( in v"431 90" sé/.
%"*— 7/1 and ò… 7/1 are in '"4"*q$%&. Since #e3 comes after #$| *T, there is a ^*%n"s"
( which is elaboration of Tot*u*q. Here, %"— is
“#e3œ‹&1 Totœ‹&' «”, ( q"T 6 and *ò(^)( is #e3.
(
Thus, these two words should be read as “*ò^-Qt1 7/1
(
%"*— 7/1”, “when *ò^-ending ( 6
%"— q"T
follows.”
(
[LSK] *òRt1 7/1 %"—T… 7/1 ^%1 7/1 '5 6/1 'Š 6/1 '5 1/1 ‘u 0 Š"T III/1 «
(
' should 6
be the substitute (no other letter, i.e., Q4€"%) (
in the place of ' of (
' ƒ^ 0y,
( 6 ending with ' ƒ^
when %"— q"T ( 0y follows.
( *ŸT^ q"T
%"—T… is 7/1 of %"— + ( *6 4oË2.1
20
No q"T*6 4oË2 #e3 is added.
226
‹¢ *+5
[LSK] l"» 1/1

( -
( %"— +
'+ ( *ò• ^ (
( %"— (
'+ 6
6.1.67 u1%^|´Š « After Q4R+i9^, (
only u remains. ( also elided.
That u is
( Q‡uy T, ( Qq"T,6 and Q#e35, it is considered to be #"*T^*o$ by 1.2.45

Since ' is
Q‡uy oq"T%6 #e35 #"*T^*o$' «(
(
Being #"*T^*o$, ^6 #e3 ( « ~ õ"T"*T^*o$"T (
is suffixed by 4.1.2 €…— …
(
Being a ^6 -ending ( 1.4.14. *â™t
entity, it is qualified as ô' by 6 (
ô' «.
Now, ô"t '$"% followed by ‹i ( should be replaced by Q4€"%
6 6
8.3.23 '9›4€"%5«. This
is negated by:
( %"— (
'+ 6
Q4€"% 6
is #"â by 8.2.23 '9›4€"%5« but negated by
8.3.25 '9 %"*— '5 ò… « ~ '5
l"s ( 8.2.36 Ôš ÷— —
| '|—3—%"— "—•2" s5« ~ ð*i ôŠ Qt1 S
l"¸ ( 8.2.39 ði" —29›t1 « ~ ôŠ
Note that 8.2.1 ^ku`y "* }'«( is used with reference to one place. Here, 8.3.25 '9 %"*— '5
ò… « is applied on the ', ( while 8.2.36 Ôš ÷— —
| '|—3—%"— "—•2" s5« is applied on the ô"t
—. ( Because these two sūtras are not modifying the same place, 8.2.1 ^ku`y "* }'«( is not
applicable for these two sūtras. The same thing is said to the next step.
l"» ( 8.4.56 u"›u "4 1 « ~ S% (ði"' (
S•y-Qn"u-^“1, when optional S•y by 8.4.56 u"›u "4 1 « is not used,
l"¸ (
227
[*u*q `k ']( 8.3.26 ‹1 '^%1 u" « ~ '5 '5 *‹T"3"' (

( optionally replaced by ', ( when followed by m'. (
' is
6 7
'( ‹ (+ ' (
1
Optional
'(
(
‹1 7/1 '^%1 7/1 u" 0 « ~ '5 1/1 '5 6/1 *‹T"3"' 7/1
• ‹1 7/1 #"*T^*o$ is ‹ (; Q after ‹ ( is for pronunciation, ƒí"%&"‡5y ; in ^% â'/.
• '^%1 7/1 '5 ^%5 3¼"T ( 5 '^%5 (115B), T*¼4; ( adjective to ‹1; in ^% â'/.
• 6
u" 0 – This is optional. The other option is Q4€"%5 6
by 8.3.23 '9›4€"%5«.
• ( the wo12, and the Q after ' is
'5 1/1 – This is wo12; ' is ( for pronunciation, ƒí"%&"‡5y .
• '5 6/1 – #"*T^*o$ is '; ( in v"431 90" sé/.
[LSK] '^%1 7/1 ‹$"%1 7/1 ^%1 7/1 'Š 6/1 '5 1/1 u" 0 «
( the substitute in the place of ' when
' is ( it is followed by '^% ‹$"%.
(
[LSK] *$' Ûi3*T; *$ Ûi3*T - Does it cause to move?
Ûi3*T = Ûi ( Si4 1 (1P) to go + *&S21( (causal) + i» (/$Ty*%/III/1
(
*$' Ûi3*T « ô"t '$"% followed by ‹i ( should be replaced
6
by Q4€"% 6
8.3.23 '9›4€"%5«. This is negated by:
(
*$' Ûi3*T 8.3.26 ‹1 '^%1 u" «
'$"%"o12"n"u^“1, when the option is not used,
*$ Ûi3*T 6
8.3.23 '9›4€"%5 «
21
( .6 (See q"T$
Ûi ( is categorized as a *'T q"T 6 925 under “Classification of roots”.) Because of 6.4.92 *'T"
(
•€5« ~ &…, *&S does not cause ƒ^q" u|*}, instead, •€ is the wo12.
228
‹¢ *+5
(u"*Ty$')( 3ui^%1 3ui" u" « ~ '5 *‹T"3"' (

• ( u/
( optionally replaced by 3/
' is ( /i
• ( i• ,( when followed by ‹ ( + 3/u ( .(
6 7
'( ( /i
‹ ( + 3/u ( (
1
optional
• ( u/
3/ • ( i• (
(
3ui^%1 7/1 3ui"5 1/1 u" 0 « ~ ‹1 7/1 '5 6/1 *‹T"3"' 7/1
• 3ui^%1 7/1 3š uš iš 3ui"5 (ID), 3ui"5 ^%"5 ^%1 3¼"T ( 5 3ui^%5 (115B), T*¼4; ( adjective to
‹1; in ^% â'/; Q after 3, ( u, ( i ( is for pronunciation, ƒí"%&"‡5y
• 3ui"5 1/1 – 3š uš iš 3ui"5 (ID)«; Q after 3, ( u, ( i ( is for ƒí"%&; this is wo12.
• 6
by 8.3.23 '9›4€"%5«.
• ‹1 7/1 #"*T^*o$ is ‹ (; Q after ‹ ( is for ƒí"%&; in ^% â'/.
• '5 6/1 – #"*T^*o$ is '; ( in v"431 90" sé/.
( *$ ã5 « Was it yesterday?
[LSK] *$3 • ã5,
(
*$' ã5 ô"t '$"% followed by ‹i ( should be replaced
6
by Q4€"% 6
(
*$3 • ã5 (u"þ) 3ui^%1 3ui" u" «
with the help of 1.3.10 3‡" û'4o16 25 '"4"'«,( Q44"*
6 ( chosen as wo12.
$ 3 • is
^“1,
*$ ã5 6
8.3.23 '9›4€"%5 «
229
(
[LSK] *$u • ;i3*T, *$ ;i3*T « Does it cause to move?
;i ( Si4 1 (1P) to go + *&S22( + i» (/$Ty*%/III/1
(
*$' ;i3*T
(
*$u • ;i3*T (u"þ) 3ui^%1 3ui" u" «
6 ( chosen as wo12.
$ u • is
^“1,
*$ ;i3*T 6
8.3.23 '9›4€"%5 «
[LSK] *$i• ( n"o3*T, *$ n"o3*T- Does it cause to be pleased?

( i» (/$Ty*%/III/1
6 (1A) to be glad + *&S +
n"o ( Ó1
(
*$' n"o3*T
*$ i• ( n"o3*T (u"þ) 3ui^%1 3ui" u" «
6 $ i• ( is chosen as wo12.
^“1,
*$ n"o3*T 6
8.3.23 '9›4€"%5 «
22
( .6 (See q"T$925
;i ( is categorized as a *'T q"T 6 under “Classification of roots”.) Because of 6.4.92 *'T" •€5«
(
~ &…, *&S does not cause ƒ^q" u|*}, instead, •€ is the wo12.
230
‹¢ *+5
[*u*q `k ']( 8.3.27 4^%1 45 « ~ ‹1 u" '5 *‹T"3"' (

( optionally replaced by 4, ( when followed by ‹ (4. (
' is
6 7
'( ‹ (+ 4 (
1
Optional
4(
(
4^%1 7/1 45 1/1 « ~ ‹1 7/1 u" 0 '5 6/1 *‹T"3"' 7/1
• 4^%1 7/1 45 ^%5 3¼"T ( 5 4^%5 (115B), T*¼4; ( adjective to ‹1; Q after 4 is
( for ƒí"%&; in ^% â'/.
• ( the wo12, and the Q after 4 is

45 1/1 – This is wo12; 4 is ( for ƒí"%&.
• ‹1 7/1 #"*T^*o$ is ‹ (; Q after ‹ ( is for ƒí"%&; in ^% â'/.

• 6
by 8.3.23 '9›4€"%5«.
• '5 6/1 – #"*T^*o$ is '; ( in v"431 90" sé/.
[LSK] 4^%1 7/1 ‹$"%1 7/1 'Š 6/1 45 1/1 u" 0 «

( the substitute in the place of ' when
4 is ( it is followed by 4^% ‹$"%.
[LSK] *$4 o( 6T 1 *$ o6T 1 - Does it take away?

o6T 1 = o6™ ( (2A) to take away + i» (/$Ty*%/III/1
*$' o( 6T 1 « ô"t '$"% followed by ‹i ( should be replaced
6
by Q4€"% 6
*$4 o( 6T 1 8.3.27 4^%1 45 « ~ ‹1 '5 u"
4$"%"o12"n"u^“1, when the option is not used,
*$ o6T 1 6
8.3.23 '9›4€"%5 «
231
‹i-( *+5 6) w0'5
[^*%n"s" `k ']( 1.1.46 w±t… »*$T… «

(
*»T w0' (
becomes w*o-Qu3u, while *$T w0' becomes Qt-Qu3u of the entity to
which w0' is enjoined.
1 1
6 6
(
*»T-w0' (
*$T-w0'
w*o-Qt… 1/2 »-*$T… 1/2 «

• w*o-Qt… 1/2 w*o5 S Qt5 S w±t… (ID).
• »-*$T… 1/2 »5 S $ ( S »$… (ID), »$… pT… 3395 T… »*$T… (116B)«. Q after » ( is ƒí"%&"‡ y.
(
[LSK] *»T-*$T… 1/2
3Š 6/1 ƒ´… 1/2 TŠ 6/1 ¨'"T 0( w*o-Qt-Qu3u… 1/2 Æ5 III/2 -
(
For which entity *»T and (
*$T w0' are enjoined, they become the beginning part and
ending part, respectively, of that entity.
“¨'"T”( word came from 1.3.10 3‡" ™¤ (
( '4o16 25 '"4"' «.
Side note:
w0' does not have its own existence. It becomes a part of the entity to which it is
(
enjoined. Because of this, if 4»• 6 (-w0' is enjoined to w'-#e3, the #e3 becomes 4$"%"*o-#e3,
not Q—"*o#e3 anymore. This is stated by this ^*%n"s" “3o"0'"Æ &/nk
6 T"Æp‹&14 0|ãt1 «”.
6th case is usually used to specify which entity the w0' should be attached to.
w0's are three types:
1. *»T –( becomes a beginning part (w*o-Qu3u) by 1.1.46 w±t… »*$T… «
2. *$T –( becomes an ending-part (Qt-Qu3u) by 1.1.46 w±t… »*$T… «
3. *'T –( comes after the last vowel by 1.1.47 *'oS9›•"T ^%5«
(
232
‹¢ *+5
[*u*q `k ']( 6 ( 2*% « ~ u" *‹T"3"' (

8.3.28 &95 $6q$
$ ( and » ( optionally becomes the end-part of ™ ( and &, ( respectively, when ™ ( and & are
(
followed by 2%.(
6 7
( &(
™/ 2% (
1
Optional
( »6$ ( w0'
$6$/
(Qt-Qu3u)
6 ( 2*% « ~ u" *‹T"3"' (

&95 6/2 $6q$ 1/1 7/1 0 7/1

• (
&95 6/2 – ™ ( S & S &… (ID), T395«; in v"431 90" sé/.
• 6 ( 1/1 – $6$ ( S »6$ ( S $6q$
$6q$ 6 ( (SD); this is w0'; being *$T, ( they become Qt-Qu3u of ™ (
( 1.1.46 w±t… »*$T…«; ƒ before $ ( is ƒí"%&"‡ y, not pT; ( thus it does not bring u&ys.
and & by
• 2*% 7/1 #e"‹"%5 2%;( in ^% â'/.
[LSK] u" 0 Æ5 III/2 «

[LSK *»å&/ #] ™$"%-&$"%395 6/2 ¨'1& 0 $6q$ 1/2 1/2
6 … w0'… Æ5
III/2
u" 0 2*% 7/1 ^%1 7/1, pe‡5y «
$6$ ( and »6$ ( are the w0', attached at the end, for ™ ( and &, ( respectively, when 2% (is
following. This operation is optional.
Before seeing the example, a u"*Ty$ related to this topic is introduced.
233
(u"*Ty$')( S39 *ÒT/3"5 2*% ^…ï% "o1*%*T u"Î' «(

( st of the class) should become 2nd of the
In the opinion of wS"3ys such as ^…ï% , S3 (1
class when followed by 2% ((sibilants). This should be said.
6 7
S3 ( 2% (
1
optional
2nd of the class
( «
S35 6/1 *ÒT/3"5 1/3 2*% 7/1 ^…ï% "o15 6/1 p*T 0 u"À' 1/1
• S35 6/1 – #e"‹"%5 S3; ( letters in the 1st of the class; in v"431 90" sé/.
• *ÒT/3"5 1/1 – letters in the 2nd of the class; this is wo12.
• 2*% 7/1 #e"‹"%5 2%;( in ^% â'/.
• ^…ï% "o15 6/1 – ^…ï% 5 w*o5 3Š 5 ^…ï% "*o5 (116B), a group of grammarians; in
¬+sé/ to “'T1”; “in the opinion of the grammarians such as ^…ï% ”.
• p*T 0 – thus.
• (
u"Î' 1/1 (
– uS (2P) (
to speak + Á3T ($'y
*&, necessity) = that which should be said.
It should be said that letters in the 2nd of the class are the substitute in the place of
S3, ( when followed by 2%,( according to ^…ï% "*o grammarians.
This u"*Ty$ is not only for $6$ ( »6$ ( w0's, but also for anywhere else applicable.
234
‹¢ *+5
6 ( 2*% « and (u"þ) S39 *ÒT/3"5 2*% ^…ïi "o1*%*T u"Î' «( are both
Because 8.3.28 &95 $6q$
optional, there can be three forms.
First, in the case of ™ ( followed by 2%.(
[LSK] #"™Ó (
( sé5, ( é5, #"™s( é5 «
#"™“
#"™23
( + sé5
#"™ ( $ ( + sé5 6 ( 2*% « ~ u"

8.3.28 &95 $6q$
( sé5
#"™ ( Ó + (u"þ) S39 *ÒT/3"5 2*% ^…ïi "o1*%*T u"Î' «(
(
( sé5
#"™Ó
When the u"*Ty$ is not used,
#"™ ( + sé5
#"™ ( $ ( + sé5 6 ( 2*% « ~ u"
8.3.28 &95 $6q$
( é5
#"™“
6 ( 2*% « is not used,

When the sūtra 8.3.28 &95 $6q$
#"™ ( + sé5
#"™s( é5
In the same manner, there are also three forms in the case of » ( followed by 2%.(
6 l ( sé5, 0Á
[LSK] 0Á 6 » ( sé5, 0& (
6 sé5 -
6 +
0& ( sé5
6 »( ( + sé5 8.3.28 &95 $6q$

0& 6 ( 2*% « ~ u"
6 l( ( + sé5 (u"þ) S39 *ÒT/3"5 2*% ^…ïi "o1*%*T u"Î' «(
0&
6 ’sé5
0Á
6 ( 2*% « is not
6 » ( sé5. When the sūtra 8.3.28 &95 $6q$
When the u"*Ty$ is not used, 0Á
(
6 sé5
used, 0&
23
( #"S (
# + QÍ • 6 0*T^k—4395 (1P) here, in the sense of going + *ò4 =
( #"Í (
Or, # + QÍ • 6 0*T^k—4395 (1P) here, in the sense of worship + *ò4 =
Either way, declined form in 1/1 is #"™(.
235
[*u*q `k ']( 8.3.29 ¸5 * q»6 ( « ~ u" *‹T"3"' (

( optionally the w0' to the
When ¸ ( is followed by , ( q is ( w*o-Qu3u.
as
5 7
¸( (
1
Optional
q»6 ( w0'
(w*o-Qu3u)
¸5 5/1
* 7/1 1/1
q»6 ( « ~ u" 0
*‹T"3"' (
7/1

• ¸5 5/1 – #"*T^*o$ is ¸ (; Q after ¸ ( is ƒí"%&"‡ y; in ^ku^y Í'/.
• * 7/1 – #"*T^*o$ is ; ( in ^% â'/.
• q»6 ( 1/1 – This is w0'; being *»T, ( it becomes w*o-Qu3u by 1.1.46 w±t… »*$T…«; ƒ before » (
is ƒí"%&"‡ y.
• u" 0 – From 8.3.26 ‹1 '^%1 u" «; this is optional.
( ^%Š 6/1 Š 6/1 q»6 1/1

[LSK] ¸"T 5/1 ( u" 0 «
q»6 ( is the optional augment (w0') attached to the beginning of (
which is preceded
by ¸ (.
q»6 ( being *»T, ( by ^*%n"s" `k 1.1.46 w±t… »*$T… «, it is attached to the beginning of . (
In the u|*§, * in the 7th case is converted into 6th case. Here is the reason:
In this sūtra, there is no 6th case-ending word. That means, “q»6 ( w0'5 $Š Qu3u5?
(Of which entity should q»6 ( w0' be a part?)” is not answered. In such a situation, there is
( which means “when there are both (5th and 7th) case
^*%n"s" “ƒn3*4oË2 1 ^Í'/*4oË29 R*i3"4«“,
endings, 5th case is stronger”. Following this ^*%n"s", the w0' becomes a part of what is
after 5th case-ending word.
y « and 1.1.67
This is based on the number of the sūtras, 1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
6
T¼"*oe§%Š«. One sūtra is ^ku y to another, which is ^%, in number. Since the 5th case is
236
‹¢ *+5
( (when two sūtras are applicable in

interpreted in the ^% `k , by ^*%n"s" 1.4.2 *u#*Ts1q 1 ^% $"3y' «,
one place at the same time, the effect which has latter number is taken), the interpretation
of 5th case takes precedence.
In our case, what is after ¸ ( is . ( Thus, q»6 ( w0' is attached to . ( Because of this, * in
the 7th case is converted into 6th case in the u|*§.
[LSK] s» („t5, s» ( t5 - Six saints.

( t5
ss +
s¸ ( + t5 8.2.39 ði" —29›t1 «
s¸ ( + q»• 6 ( t5 8.3.29 ¸5 * q»6 ( « ~ u"
s¸ ( + T ( t5 8.4.55 Ó*% S « ~ ði" S% (
s» ( + T ( t5 8.4.55 Ó*% S « ~ ði" S% (
s» („t5
^“1
( t5
ss +
s¸ ( + t5 8.2.39 ði" —29›t1 «
s» ( + t5 8.4.55 Ó*% S « ~ ði" S% (
srt5
Note:
7/1
One may doubt that instead of * , 5 6/1 could have been said to achieve the same
result. The answer is: by saying * , half a '"`" is saved, compared to 5. There is a saying
6
“Qqy'"`"i"•u14 ^`9„u 'Õ"t1 u•3"$%&"5« (by shortening the length by half '"`", grammarians
consider it as a celebration like for the birth of a son.)“
237
[*u*q `k ']( 8.3.30 4š « ~ * q»6 ( u" *‹T"3"' (

( followed by , ( q is
When 4 is ( optionally the w0' to the ( w*o-Qu3u.
as
5 7
4( (
1
Optional
q»6 ( w0'
(w*o-Qu3u)
45 5/1 S 0 « ~ * 7/1
q»6 1/1 (
( u" 0 *‹T"3"' 7/1
• 45 5/1 – #"*T^*o$ is 4; ( Q after 4 is
( ƒí"%&"‡ y; in ^ku^y Í'/.
• S 0 – Connects to the previous sūtra.

• * 7/1 – #"*T^*o$ is ; ( in ^% â'/.
• q»6 1/1
( – this is w0'; being *»T, ( it becomes w*o-Qu3u by 1.1.46 w±t… »*$T…«.
• u" 0 – From 8.3.26 ‹1 '^%1 u" «; this is optional.
( ^%Š 6/1 Š 6/1 q»6 1/1

[LSK] 4-Qt"T 5/1 ( u" 0 «
q»6 ( is optionally the augment (w0') attached to the beginning of (
which is
(
preceded by 4-ending ô.
In i•*6 }"t$…'o6 /u|*§, “ôŠ” is taken from Q*q$"% `k 8.1.16 ôŠ «, and using Tot*u*q.
This interpretation does not contribute any plus or minus.
[LSK] s„5, 4 ( 5- Being, he.

( 5
4+
( q»6 ( 5
4+ 8.3.30 4š « ~ * q»6 ( with the help of 1.1.46 w±t… »*$T… «
( T( 5
4+ 8.4.55 Ó*% S « ~ ði" S% (
s„5
^“1
t5
238
‹¢ *+5
[*u*q `k ']( 6 ( « ~ 45 ôŠ u" *‹T"3"' (

8.3.31 *2 T$
( the end of ô is followed by 2, ( T$
When 4 at ( Qt-
6 ( is optionally the w0' to the 4 as
Qu3u.
6 7
ô 4( 2(
1
Optional
6 ( w0'
T$
(Qt-Qu3u)
6 ( 1/1 « ~ 45 6/1 ôŠ 6/1 u" 0 *‹T"3"' 7/1

*2 7/1 T$ (

• *2 7/1 – #"*T^*o$ is 2; ( in ^% â'/.
• 6 ( 1/1 – This is w0'; being *$T, ( it becomes Qt-Qu3u by 1.1.46 w±t… »*$T…«; ƒ before
T$
$ ( is ƒí"%&"‡ y.
• 45 6/1 – From 8.3.31 4š«; by *un*´*u^*%&"', it becomes 6th case-ending; adjective to
ôŠ; by Tot*u*q, final meaning is “4"tŠ”.
• ôŠ 6/1 – From 8.1.16 ôŠ«; in v"431 90" sé/.
• u" 0 –From 8.3.26 ‹1 '^%1 u" «. This is optional.
6 ( 1/1 u" 0 «
[LSK] ô"tŠ 6/1 4Š 6/1 21 7/1 ^%1 7/1 T$
( the end of ô, which is followed
6 ( is optionally the w0' attached to the end of 4 at
T$
by 2. (
In the u|*§, instead of “4"tŠ ôŠ”, “ô"tŠ 4Š” is told. However, it does not make
6 ( becomes the Qt-Qu3u (the last part) of the ô with the help of
any difference because T$
1.1.46 w±t… »*$T…«.
239
[LSK] uf5,6 Ø•f56 , Øˆ2f56 , vf56 -

The first form is made by applying all the possible sūtras.
( 2f56
4+
4 T( $
6 ( + 2f56 6 ( « ~ ôŠ 4Š u"
8.3.31 *2 T$
( +
4S ( 2f56 6 š56 «
8.4.40 Æ95 š4"
ÈS ( 2f56
( + 6 š56 «
8.4.40 Æ95 š4"
ÈS ( ¶f56
( + 8.4.63 2Ÿ¶9›*» « ~ u"
( ¶f56
È+ 8.4.65 ð%9 ð*% u&Ë « ~ i9^5 ‹i5 u"
uf56
The second form is without the optional ð% (i9^ by 8.4.65 ð%9 ð*% u&Ë «.
( 2f56
4+
4 T( $
6 ( + 2f56 6 ( « ~ ôŠ 4Š u"
8.3.31 *2 T$
( +
4S ( 2f56 6 š56 «
8.4.40 Æ95 š4"
ÈS ( 2f56
( + 6 š56 «
8.4.40 Æ95 š4"
ÈS ( ¶f56
( + 8.4.63 2Ÿ¶9›*» « ~ u"
Ø•f56
The third form is without the optional ¶• 8.4.63 2Ÿ¶9›*» «.
( 2f56
4+
4 T( $
6 ( + 2f56 6 ( « ~ ôŠ 4Š u"
8.3.31 *2 T$
( +
4S ( 2f56 6 š56 «
8.4.40 Æ95 š4"
ÈS ( 2f56
( + 6 š56 «
8.4.40 Æ95 š4"
Øˆ2f56
6 ( w0' by 8.3.31 *2 T$
The fourth form is without the optional T$ 6 ( «.
( 2f56
4+
( 2f56
È+ 6 š56 «
8.4.40 Æ95 š4"
vf56
240
‹¢ *+5
[*u*q `k ']( (
6 *4e'
8.3.32 ™'9 •€"o*S ™'& «( ~ ô"T ( *‹T"3"' (
™,( &, ( and 4 will
( (
become the beginning part of QS which is preceded by ô which
ends with ™,( &, ( and 4, ( respectively, which are preceded by •€.
5 5 7
ô ™' (
•€ QS (
1
™'»6 ( w0'
(w*o-Qu3u)
( Q*S 7/1 ™'»6 1/1

™'5 5/1 •€"T 5/1 ( *4e' 0( « ~ ô"T 5/1
( (
*‹T"3"' 7/1
• ™'5 5/1 – #e"‹"%5 ™', ( contains three letters, i.e., ™,( &, ( 4; ( being an adjective to ô"T, ( by
(
Tot*u*q, final meaning is “™'-Qt"T (
ô"T”.(
• •€"T (5/1 – short vowel; in ^ku^y Í'/.
• Q*S 7/1 – #e"‹"%5 QS; ( in ^% â'/.
• ™'»6 1/1
( (
– This is w0'; #e"‹"%5 ™' with ƒ for ƒí"%& and » ( as pT; ( hence three w0's are
understood, i.e., ™6 » (, &»6 (, and 4»6 (; being *»T, ( it becomes w*o-Qu3u by 1.1.46 w±t… »*$T…«.
• *4e' ( 0 – Here, it means #"35, most of the time. This does not mean “optional”. The
absence of ™'»6 ( is seen only by *2U#390, such as 1.4.14 *â™t
6 ô' «( and 3.1.32 4"±t"
q"Tu5«.
• (
ô"T 5/1 – From 8.1.16 ôŠ«; by *un*´*u^*%&"', it becomes 5th case-ending; in ^ku^y Í'/.
( ^%5 1/1 35 1/1 ™' 1/1

[LSK] •€"T 5/1 ( Tot' 1/1
( 3T 1/1
( ô' 1/1
( T¼"T 5/1
( ^%Š 6/1 QS5 6/1 ™'»6 1/1
( «
™'»6 is (
( w0' attached to the beginning of QS which is after the ô which ends ™' (
preceded by short vowel.
241
(
Three examples are given as the #e"‹"% ™' has three letters.
Note that only one form is given because this rule is *4e, always applied and only
*2U have any choice in this.
( "ö" 1/1 «
[LSK] #e™™
#e™ ( + wö"
#e™ ( + ™ ( wö" (
6 *4e'
8.3.32 ™'9 •€"o*S ™'& «( ~ ô"T (
(
By 1.1.72 314 *u*qÆotŠ«, “Qt” is added to make “™'-Qt"T ”(
6
1.1.66 T¼"*oe§%Š« (
for •€"T “^%5”, (
and ™'t"T ô"T (
“^%Š”
1.3.10 3‡" ¤'4o16 25 '"4"'«( for connecting ™'5 and ™'»6 (
Because there is no 6th case-ending word, ƒn3*4oË2 1 ^*Í'/*4oË29 Ri/3"4«(
to change the 7th case of Q*S to 6th case.
By 1.1.46 w±t… »*$T…«, ™'»6 ( is w*oQu3u.
1/1
6
[LSK] 0Á&/25 « üÎT6 5 1/1 -
6 +
Original forms are 0& ( 25 and 4 +
( QÎT6 5.
( wö*4” and “Q*¼4 +

Note: In the case of Q*¼ü"ö*4, it can be resolved into “Q*¼4 + (
4 + wö*4” because of this sūtra.
242
‹¢ *+5
‹i-( *+5 7) 2nd 7• •-#$%&' (
Now, a new section called “2nd 7• •-#$%&” starts. The “1st 7• •-#$%&” is from 8.2.66
6 7• :«. This 2nd section starts from Q*q$"% `k 8.3.1 'Tu6 9 7•
—s9 ¬}6 … ¶¡* «, which runs
6
until 8.3.12. 7• is the wo12 in this section. Sūtras 8.3.2 Q`"44"* y T 6 u" « and 8.3.4
$ ^kuŠ
6
Q44"* ( Q4€"%5
$"T ^%5 6 6 «
« teach how the 7• wo12 should be applied. Sūtras from 8.3.5 '5 *»
to 8.3.12 $"4"l1*¸T1 « teach in which occasions this 7• • takes place.
[*u*q `k ']( 6 « ~ 7• : *‹T"3"' (

8.3.5 '5 *»
( replaced by 7• when it is followed by »6 ( w0'.
The last letter of ' is
6 7
of ' ( ' ( »6 (
1
7•
'5 6/1 *» (
6 7/1 « ~ 7• : 1/1 *‹T"3"' 7/1

• '5 6/1 – #*T^*o$ is '; ( in v"431 90" sé/.
• 6 7/1 – w0' given in the section starting from 6.1.135 »6 ( $"T ^k( u5y «; in ^% â'/.
*»
• 7• : 1/1 – This is wo12; from Q*q$"% `k 8.3.1 'Tu6 9 7• ¬}6 … ¶¡* «.
[LSK] '5 6/1 7• : 1/1 *»

6 7/1 «
(
7• is the substitute in the place of the last letter of ' when it is followed by »6 (.
^*%n"s" `k 1.1.52 Qi9›•Š « is required.
( $T"y
'+
( »6 ( + $T"y
'+ 1 5 $%9T… nks&1 « ~ »6 ( $"T ^k( u5y
6.1.137 †3^6y 8
7• + ( $T"y
+ 6 « ~ 7• :, with the help of 1.1.52 Qi9›•Š «
8.3.5 '5 *»
The following two sūtras teach how the 7• • should take place in this section.
243
[*u*q `k ']( 6 $5 ^kuŠ

8.3.2 Q`"44"* y T 6 u" « ~ *‹T"3"' (
In this 7• • section, the preceding letter of 7• optionally becomes nasalized.
6
u&y 7•
1
6
Q44"* $
Optional
Q` 0 Q44"*
6 $5 1/1 ^kuŠ (
y 6/1 T 6 0 u" 0 « ~ *‹T"3"' 7/1
• Q` 0 – In this 7• •-Q*q$"%; differentiating from another 7• •#$%& starting from 8.2.66
6 7• :.
—s9
• 6
Q44"* 6
$5 1/1 – This is wo12; Q44"* $ is a !" defined by 1.1.8 'Ó 6
6 4"* $"uS49›44"* $5«;
6
by v"4›1 t%T'5«, the closest Q44"* ( chosen; in other words, v"*44 (
$ letter to the v"*44 is
becomes nasalized.
• y 6/1 – 7• •"T ^k( uŠ
^kuŠ y u&yŠ is understood; in v"431 90" sé/.
• T 6 0 – ÐU"‡ y', ( for the sake of clarity; while 7• • is *4e, compulsory, this Q44"*
6 $• is
optional; to show this inequality, u•sÅ, T 6 is mentioned.
[LSK] Q` 0 7• •#$%&1 7/1 %95 5/1 ^kuŠ

y 6/1 Q44"*
6 $5 1/1 u" 0 «
6
In this 7• • section, Q44"* $ letter is optionally the substitute in the place of the
previous letter of 7• .
( $T"y
'+
( »6 ( + $T"y
'+ 6.1.137 †3^6y 8
1 5 $%9T… nks&1 «
7• + ( $T"y
8.3.5 '5 *»
• % (+ ( $T"y
+ 6
8.3.2 Q`"44"* y T 6 u"«
$5 ^kuŠ
8.2.1 ^ku`y "* }'«( is not applicable here because 8.3.2 and 8.3.4 are considered to be
Q4u|6 *§ for the sūtras that follow.
244
‹¢ *+5
[*u*q `k ']( 8.3.4 Q44"* (

6 $"T ^%9›4 6
€"%5 « ~ %95 ^ku"yT ( *‹T"3"' (
6
Other than Q44"* 6
$, (when Q44"* 6
$ option is not taken,) Q4€"% is augmented after
the letter before the 7• .
( ^%5
y "T 5/1
^ku¼
1
6
Q4€"% w0'
u&y 7•
Optional
%95 5/1 ^ku5y
6
Q44"* ( ^%5 1/1 Q4€"%5
$"T 5/1 6 1/1
« ~ %95 5/1 ^ku¼ (
y "T 5/1 (
*‹T"3"' 7/1
• 6
Q44"* (
$"T 5/1 6
– “Q44"* $ *u‹"3, giving up the nasal” is understood; this usage of 5th
case is by (u"þ) ^Í'/*uq"4 1 Þw9^1 $'yÁ3^6 û"4'”,( instead of stating in 2nd case with Þ^-(
(
ending word, the same thing can be told in 5th case; examples are: #" "o' w7ã ^Ÿ3*T «
(
= #" "o"T ^Ÿ3*T «.
• ^%5 1/1 – This shows the position of w0'.
• 6
Q4€"%5 1/1 – This is w0'.
• y "T. (
%95 5/1 – 7• which was given by 8.3.5 to 8.3.12; in *oÏ3901 ^Í'/ related to ^ku¼
• ^ku¼ (
y "T 5/1 – before 7• ; in *oÏ3901 ^Í'/ related to ^%5.
6
[LSK] Q44"* ( *u‹"3 0 %95 5/1 ^ku¼
$' 2/1 ( ^%5 1/1 Q4€"%-w0'5
y "T 5/1 6 1/1
-
6
Leaving Q44"* 6
$ option, Q4€"% as w0' comes after the previous letter of 7• .
( $T"y
'+
( »6 ( + $T"y
'+ 6.1.137 †3^6y 8
1 5 $%9T… nks&1 «
7• + ( $T"y
8.3.5 '5 *»
% (+ ( $T"y
+ 6
8.3.4 Q44"* (
$"T ^%9›4 6
€"%5«
8.2.1 ^ku`y "* }'«( is not applicable here because 8.3.2 and 8.3.4 are considered to be
Q4u|6 *§ for the sūtras that follow.
245
[*u*q `k ']( 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ *‹T"3"' (

%1Ý at the end of ô is replaced by *u 0y when followed by Ó% (or Qu "4.
6 7
ô %( (
Ó%/Qu "4
1
*u —y4/3
(
Ó%-Qu (
"4395 7/2 *u —y4/35 1/1 « ~ %5 6/1 ôŠ 6/1 *‹T"3"' 7/1
• (
Ó%-Qu "4395 7/2 – #e"‹"%5 Ó%,( all the hard consonants; Qu "4 is !" defined as the
absence of letter, by 1.4.110 *u%"'9›u "4'«;( in ^% â'/
• *u —y4/35 1/1 – This is wo12; *u —y4/3 is synonym for *u 0y.
• %5 6/1 – From 8.3.14 %9 *%«; since it is '"4"*q$%& to ôŠ, by Tot*u*q, it is read together
as “%1Ý"tŠ ôŠ”; in v"431 90" sé/.
• ôŠ 6/1 – From Q*q$"% `k 8.1.16 ôŠ«; in v"431 90" sé/.
[LSK] Ó*% 7/1 Qu "4 1 7/1 S 0 ô"tŠ 6/1 %1ÝŠ 6/1 *u 0y5 1/1«
*u 0y is the substitute in the place of the last letter of %1Ý-ending ô when Ó% (or Qu "4
follows.
( $T"y
'+
( »6 ( + $T"y
'+ 6.1.137 †3^6y 8
1 5 $%9T… nks&1 «
7• + ( $T"y
8.3.5 '5 *»
• % (+ ( $T"y
+ 6
8.3.2 Q`"44"* y T 6 u" «
$5 ^kuŠ
:• + ( $T"y
+ 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ, with the help of 1.1.52 Qi9›•Š «
^“1
% (+ ( $T"y
+ 6
8.3.4 Q44"* (
$"T ^%9›4 6
€"%5«
:+ ( $T"y
+ 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ , with the help of 1.1.52 Qi9›•Š «
246
‹¢ *+5
It is easier to understand the next u"*Ty$, which is in the context of change of *u 0y to

, ( when the following two sūtras, which are introduced in i•*6 }"t$…'o6 / later, are studied
first. Thus, the order of sūtra is modified here for the next two sūtras.
[*u*q `k ']( 8.3.34 *u —y4/3Š 5 « ~ Ó*% *‹T"3"' (

( the substitute for *u 0y followed by Ó%.(
is
6 7
*u —y4/3 Ó% (
1
(
(
*u —y4/3Š 6/1 5 1/1 « ~ Ó*% 7/1 *‹T"3"' 7/1
• *u —y4/3Š 6/1 – *u —y4/3 is synonym for *u 0y; in v"431 90" sé/.
• 5 1/1 – This is wo12; Q after ( ƒí"%&"‡ y.
is
• Ó*% 7/1 – #e"‹"%5 Ó%,( all the hard consonants; in ^% â'/.
[LSK] Ó*% 7/1 «

( the substitute in the place of *u 0y when Ó% (follows.
is
247
The next sūtra is Qû"o of 8.3.34 *u —y4/3Š 5«, and teaches that the change is
optional when 2% (follows.
[*u*q `k ']( 8.3.36 u" 2*% « ~ *u —y4/3Š *u —y4/35 *‹T"3"' (

*u 0y is optionally the substitute for *u 0y when followed by 2%.( The other option is . (
6 7
*u —y4/3 2% (
1
*u —y4/3
Optional
(
u" 0 2*% 7/1 « ~ *u —y4/3Š 6/1 *u —y4/35 1/1 *‹T"3"' 7/1
• 2*% 7/1 – #e"‹"%5 2%,( all the sibilants; in ^% â'/.
• *u —y4/35 1/1 – This is wo12.
*u 0y is optionally the substitute in the place of *u 0y when 2% (follows. The other

option is . (
248
‹¢ *+5
This u"*Ty$ is Qû"o to 8.3.36 u" 2*% «. The change to ( *4e, compulsory for *u 0y of
is
', ( ^'6 , ( and $"4. (
(u"*Ty$')( ^$"4"
6 9 u´Z5« ~ *u —y4/3Š *‹T"3"' (
( always replaced by . (
*u 0y of ', ( ^'6 , ( $"4 is
6
*u —y4/3
of ', ( ^'6 , ( $"4 (
1
(
Compulsory
( ™
'-^ (
6 -( $"4"' 6/3 (
5 1/1 u´Z5 1/1 « ~ *u —y4/3Š 6/1 *‹T"3"' 7/1
• ( ™
'-^ (
6 -( $"4"' 6/3 ( ^'6 S
– 'S ( †™
( $"4 S 6 $ ( these are the words subject to
( "45 (ID), T1s"' «;
7• • and *u 0y in the 7• •#$%&; in ¬+sé/ to *u —y4/3Š.

is
• ( say+ TZ $'y*& with the sense of necessity; “this should have been
u´Z5 1/1 – uS to
said”.
This u"*Ty$ is not only Qû"o to 8.3.36 u" 2*% «, but any other sūtra which can bring a
letter other than ( the *u 0y of ', ( ^'6 , ( and $"4. (
for
249
[LSK] -
• T"y, ÷-T"y - The agent of decorating.
( $T"y
'+
( »6 ( + $T"y
'+ 6.1.137 †3^6y 8
1 5 $%9T… nks&1 «
7• + ( $T"y
+ 6 « ~ 7• : , with the help of 1.1.52 Qi9›•Š «
8.3.5 '5 *»
• % (+ ( $T"y
+ 6
8.3.2 Q`"44"* y T 6 u" «
$5 ^kuŠ
:• + ( $T"y
(
• + ( $T"y
+ 6
(u"þ) ^$"4" 9 u´Z5 « ~ *u —y4/3Š 5
÷• -T"y
6
Q4€"%-^“1
( $T"y
'+
( »6 ( + $T"y
'+ 6.1.137 †3^6y 8
1 5 $%9T… nks&1 «
7• + ( $T"y
8.3.5 '5 *»
% (+ ( $T"y
+ 6
8.3.4 Q44"* (
$"T ^%9›4 6
€"%5«
:+ ( $T"y
(
+ ( $T"y
+ 6
(u"þ) ^$"4" 9 u´Z5 « ~ *u —y4/3Š 5
÷-T"y
250
‹¢ *+5
In the same manner, a few more sūtras from the 7• •#$%& are introduced. The next
sūtra shows the case of ^'6 , ( related to the u"*Ty$ † 6 "4" 9 u´Z5 «.
[*u*q `k ']( 6 Ó[†%1« ~ 7• : *‹T"3"' (

8.3.6 ^'5
( replaced by 7• when it is followed by Q†% Ó3. (
The last letter of ^'6 is
6 7
of ^'6 ( '( ( Q' (
Ó3 +
1
7•
( 7/1 « 7• : 1/1 *‹T"3"' 7/1

6 6/1 Ó*3 7/1 Q'-^%1
^'5 (

• 6 6/1 – #*T^*o$ is ^'6 ; ( in v"431 90" sé/.
^'5
• Ó*3 7/1 – #e"‹"%5 Ó3, ( letters in 1st and 2nd of the classes; in ^% â'/.
• Q'-^%1 ( 3¼"T ( 5 Q†%5 (115B) = Ó3, ( T*¼4«;( adjective to Ó*3; in ^% â'/.
( 7/1 – Q' ^%5
[LSK] Q†%1 7/1 Ó*3 7/1 ^'5

6 6/1 7• : 1/1 «
(
7• is the substitute in the place of the last letter of ^'6 when it is followed by Ó3, (
( its ^%.
which has Q' as
251
6
^'"4 (
$9*$i5 6
p*T ^-9*$i5 (KD) « A male cuckoo.
( • 6 + $9*$i + • 6 2.1.57 *u21s& *u21³&
^6 + 1 Rºi' «( ~ T°7s5
6 '" 5
1.2.46 $| §*}T '" "š« ~ #"*T^*o$' (
( $9*$i
^6 + 6 q"T#6 "*T^*o$395 « ~ $ (
2.4.71 ^9
( $9*$i
^'6 + 8.2.23 390"tŠ i9^5 « ~ ôŠ
6
Q4€"% ( (^þ) *4*'§"^"31 4 •*'*§$Š"A^"35 «24
comes back to ' by
^7•6 + $9*$i 6 Ó[†%1 « ~ 7• , with the help of 1.1.52 Qi9›•Š «
8.3.6 ^'5
^ • 6 % (+ $9*$i 6
8.3.2 Q`"44"* y T 6 u" «
$5 ^kuŠ
^:• 6 + $9*$i 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ, with the help of 1.1.52 Qi9›•Š «
( $9*$i
^•6 + 6
(u"þ) ^$"4" 9 u´Z5 « ~ *u —y4/3Š 5
• 6 9*$i
^-
^“1
( $9*$i
^'6 +
^7•6 + $9*$i 6 Ó[†%1 « ~ 7•
8.3.6 ^'5
^ 6 % (+ $9*$i 6
8.3.4 Q44"* (
$"T ^%9›4 6
€"%5«
^:6 + $9*$i 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ
( $9*$i
^6 + 6
(u"þ) ^$"4" 9 u´Z5 « ~ *u —y4/3Š 5
6
^-9*$i
6
Note that the u"*Ty$ ^$"4" 9 u´Z5 « ~ *u —y4/3Š 5 not only prevents u•$*®$
*u —y4/3, but also *—;"'ki/3 told by 8.3.37 $6x95 ≍$≍ ^… S «.
24
*4*'§Š (of the cause) Q^"35 (removal) *4*'§"^"35, T*¼4 ( *T 4 •*'*§$Š (of the effect) Q*^ Q^"35 «
When the cause of removed, the effect is also gone.
252
‹¢ *+5
[*u*q `k ']( 8.3.7 4Ÿ¶Z#2"4 «( ~ Q†%1 ôŠ 7• : *‹T"3"' (

( the end of ô, excluding #2"4, ( is replaced by 7• when it is followed by Q†% ¶u. (
4 at
6 7
ô 4( ( Q' (
¶u +
1
7•
( « ~ Q'-^%1
45 6/1 ¶*u 7/1 Q#2"4 7/1 ( 7/1 ôŠ 6/1 7• : 1/1 *‹T"3"' 7/1
(

• 45 6/1 – #*T^*o$ is 4; ( since it is '"4"*q$%& to ôŠ, by Tot*u*q, it is read together as
“4$"tŠ ôŠ”; in v"4 1390" sé/.
• ( ( » ( l ( T; ( T"iZ, 'kqÕ
¶*u 7/1 – #e"‹"%5 ¶u; ( i.e., S ¶ y , and o• letters in 1st and 2nd of the
classes; in ^% â'/.
• Q#2"4 ( 6/1 – #*T^*o$ is Q#2"4; ( 4 #2"425
( Q#2"4; ( “excluding the word #2"4”;( in v"431 90"
sé/, which is â, elided.

• ( 3¼"T ( 5 Q†%5 (115B) = ¶u, ( T*¼4«;( adjective to ¶*u; ¶u is
( 7/1 – Q' ^%5
Q'-^%1 ( #e"‹"%
consisting 1st and 2nd of Su0y, »u0y, and Tu0y (S, ( ¶ (, » (, l (, T, ( ‡);( Q' is
( #e"‹"% consisting of
all the vowels, semi-vowels, and nasals; in ^% â'/.

• ôŠ 6/1 – From Q*q$"% `k 8.1.16 ôŠ«; in v"431 90" sé/.
(
[LSK] Q†%1 7/1 ¶*u 7/1 4-QtŠ 6/1 ôŠ 6/1 7• : 1/1; 4 0 T 6 0 #2"4-2²Š 6/1
-
(
7• is the substitute in the place of the last letter of 4-ending ô, but not of the word
“#2"4”,( when the ô is followed by ¶u, ( which has Q' as
( its ^%.
25
(
# + 2' ƒ^2'1 + *ò^ (
# + 2"' ( 6
6.4.15 Q44"* $Š *ò « ~ o/•5y
# + 2"4 ( 8.2.64 '9 49 q"T95 «
253
i•*6 }"t$…'o6 / introduces 8.3.34 *u —y4/3Š 5 « now. Then, it shows examples.
(
[LSK] S*¨4 `"3€ « Oh! ViJ u! Protect us.
( 4-ending
S*¨4 is ( ( ¶u, ( and % (is Q'. (
ô, followed by `, ( of which T is
S*¨7• + `"3€ 8.3.7 4Ÿ¶Z#2"4 «( ~ Q†%1 ôŠ 7•
S*¨• % (+ `"3€ 6
8.3.2 Q`"44"* y T 6 u" «
$5 ^kuŠ
S*¨• : + `"3€ 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ
S*¨• ( `"3€
+ 6
(u"þ) ^$"4" 9 u´Z5 « ~ *u —y4/3Š 5
(
S*¨• `"3€
^“1
S*¨7• + `"3€ 8.3.7 4Ÿ¶Z#2"4 «( ~ Q†%1 ôŠ 7•
S*¨ % (+ `"3€ 6
8.3.4 Q44"* (
$"T ^%9›4 6
€"%5«
S*¨: + `"3€ 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ
( `"3€
S*¨ + 6
(u"þ) ^$"4" 9 u´Z5 « ~ *u —y4/3Š 5
(
S*¨ `"3€
(
When ¶u comes in Su0y and »u0y, example will be S*¨• U/$€, S*Ü/$€, and S*¨• š%,
6 š5«
S*¨š%, respectively with 8.4.40 Æ95 š4" 6 U56 « on the . (
6 and 8.4.41 U4
ô$| e is done.
( ?( #2"4 1/1
[LSK] Q#2"4 *$' ( T49*T «
( 4-ending
#2"4 is ( ( T49*T is Q†%¶u. ( Because of the mentioning of “Q#2"4”(
ô, and T of
( 7• • does not apply.
in 8.3.7 4Ÿ¶Z#2"4 «,
[LSK] ôŠ p*T *$'?( ‹*t -
( i» (/$Ty*%/III/1
‹4 +
( *T
‹4 +
6
By 1.4.14 *â™t ( not ô"t. Thus even though
ô'«,( the ô"t is at the p of *T. The 4 is
( *T is Q†%¶u, ( and it is following, the 7• • does not apply.
T of
254
‹¢ *+5
[*u*q `k ']( ( « ~ 7• : ƒn3‡" *‹T"3"' (

8.3.10 4Q4 ^1
( 4Q is optionally replaced by 7• when it is followed by ^. (
4 of
6 7
of 4Q4 ( 4( ^(
1
optional
7•
( ^1 7/1 « ~ 7• : 1/1 ƒn3‡" 0 *‹T"3"' 7/1

4Q4 6/1 (

• (
4Q4 6/1 – #*T^*o$ is 4Q4; ( 4Q4 is
( Q4$%&,
6 mimicking of the sound of 4| 2² in 2/3; in v"431 90"
sé/, which is â.
• ( ƒí"%&"‡ y; in ^% â'/.
^1 7/1 – #*T^*o$ is ^; Q after ^ is
• ƒn3‡" 0 – From 8.3.8 ƒn3‡“«6y ; this means “in both ways”, which makes this rule
optional.
[LSK] “4Q4”( p*T 0 QŠ 6/1 7• : 1/1 u" 0 ^1 7/1 -

7• is optionally the substitute in the place of the last letter of 4Q4, ( when followed by ^. (
p*T QŠ brings back the â sé/ ‘$uS4', ( and also tells that “4Q4”( is Q4$%&,
6 an
onomatopoeia.
(
The example is “4Q4 ^"*‹, Protect the people”.
( ^"*‹
4Q4 +
4Q7• + ^"*‹ ( « 7•
8.3.10 4Q4 ^1
4Q • % (+ ^"*‹ 6
8.3.2 Q`"44"* y T 6 u" «
$5 ^kuŠ
4Q:• + ^"*‹ 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ
Now, *u 0y is followed by ^. ( The next sūtra is applicable.
255
This sūtra is Qû"o for 8.3.34 *u —y4/3Š 5«. Instead of , ( this sūtra optionally gives
*—;"'ki/3 and ƒ^:"4/3. The other option is *u 0y.
[*u*q `k ']( 8.3.37 $6x95 ≍$≍^… S « ~ *u —y4/3Š *u —y4/35 *‹T"3"' (

*u 0y is optionally replaced by *—;"'ki/3 and ƒ^:"4/3, when followed by $u0y and
û0y, respectively.
6 7
*u —y4/3 $6/^ 6
1
≍ *—;"'ki/3/
≍ ƒ^:"4/3
Optional
(
$6-x95 7/2 ≍$≍^… 1/2 S 0 « ~ *u —y4/3Š 6/1 *u —y4/35 1/1 *‹T"3"' 7/1
• $6-x95 7/2 – #"*T^*o$ is $6^;6 $65 S ^56 S $6^k (ID), T395; in ^% â'/.
• ≍$≍^… 1/2 – This is wo12. They indicate *—;"'ki/3 and ƒ^:"4/3.
• S 0 – This S brings *u 0y as the other option.
• *u —y4/35 1/1 – From 8.3.35 2^ y%1 *u —y4/35«; this is brought by the S to eliminate the
possibility for ( 2 by 8.3.34 *u —y4/3Š 5«.
wo1
( *u 0y5 1/1«
[LSK] $u0Ë 7/1 û0Ë 7/1 S 0 *u 0yŠ 6/1 ≍$≍^… 1/2 Æ5 III/2, S"T 5/1
When $u0y and û0y follow, *—;"'ki/3 and ƒ^è"4/3, respectively, are the substitute in
the place of *u 0y optionally. The S in the sūtra indicates the other option is *u 0y.
256
‹¢ *+5
[LSK] 4Q≍ (
• ^"*‹, 4Q≍ ^"*‹, 4Q: ^"*‹, « 4Q4 ^"*‹ -
Though i•*6 }"t$…'o6 / shows only four forms, there should be five forms:
6
1) 7• • with Q44"* $, then ƒ^:"4/3 4Q≍
• ^"*‹
6
2) 7• • with Q44"* $, then *u 0y 4Q:• ^"*‹
6
3) 7• • with Q4€"%, then ƒ^:"4/3 4Q≍ ^"*‹
6
4) 7• • with Q4€"%, then *u 0y 4Q: ^"*‹
5) Without 7• • (
4Q4 ^"*‹
( ^"*‹
4Q4 +
4Q7• + ^"*‹ ( « ~ 7•
8.3.10 4Q4 ^1
4Q • % (+ ^"*‹ 6
8.3.2 Q`"44"* y T 6 u" «
$5 ^kuŠ
2) 4Q:• + ^"*‹ 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ
1) 4Q≍
• + ^"*‹ 8.3.37 $6x95 ≍$≍^… S «
5) ( ^"*‹
4Q4 +
4Q7• + ^"*‹ ( « ~ 7•
8.3.10 4Q4 ^1
4Q % (+ ^"*‹ 6
8.3.4 Q44"* (
$"T ^%9›4 6
€"%5«
4) 4Q: + ^"*‹ 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ
3) 4Q ≍ + ^"*‹ 8.3.37 $6x95 ≍$≍^… S «
257
The next sūtra gives wl1*¸T !" which is used in the following 7• •*u*q.
There is a section called *Ò7*´-#$%&, starting from the sūtra 8.1.1 uyŠ Ò1 «, which
teaches that a whole word is said twice under certain conditions specified by the following
sūtras in the section. The sūtra 8.1.2 TŠ ^%'"l1*¸T' «( gives wl1*¸T !" to the latter one of the
repeated words.
[ !" `k ']( 8.1.2 TŠ ^%'"l1*¸T' «(

Of the *Ò7´, the latter one is termed wl1*¸T.
( wl1*¸T' 1/1
TŠ 6/1 ^%' 1/1 ( «

• TŠ 6/1 – Here, the topic, Q*q$"%, is *Ò7*´, repetition of a word, by Q*q$"% `k 8.1.1 uyŠ
Ò1«; thus TŠ is understood as *Ò7´Š, of what is repeated; in ¬+1 sé/ to ^%'. (
• (
^%' 1/1 – This is !/; of the two words as *Ò7*´, the first one is ^ku,y the second one is ^%.
• (
wl1*¸T' 1/1 – This is !".
( wl1*¸T' 1/1
[LSK] *Ò7´Š 6/1 ^%' 1/1 ( Š"T III/1
( -
Of the repeated words, the latter one is termed wl1*¸T.
258
‹¢ *+5
[*u*q `k ']( 8.3.12 $"4"l1*¸T1 « ~ 7• : *‹T"3"' (

( $"4 which
4 of ( is followed by wl1*¸T is replaced by 7• .
6 7
$"4 (
of $"4 ( 4(
(wl1*¸T)
1
7•
( wl1*¸T1 7/1 « ~ 7• : 1/1 *‹T"3"' 7/1

$"4 6/1 (

• $"4 ( 6/1 – #*T^*o$ is $"4; ( $"4 is
( Q4$%&,
6 (
mimicking of the sound of *$' 2² in 2/3; in
v"431 90" sé/, which is â.
• (
wl1*¸T1 7/1 – The $"4 must have been repeated by any injunction in the *Ò7*´ section;
the latter one of the repeated $"4; ( in ^% â'/.
( 4$"%Š 6/1 7• : 1/1 Š"T III/1

[LSK] $"4 6/1 ( wl1*¸T1 7/1 «
(
7• is the substitute in the place of 4 (the last letter) of “$"4”,( when followed by wl1*¸T.
( $"4 (
$"4 + 8.1.4 *4eu/`395 « ~ uyŠ Ò1
When constancy or repetition is intended, the whole word is said twice.
$"7• + $"4 ( 8.3.12 $"4"l1*¸T1 « ~ 7• with 1.1.52 Qi9›•Š «
$"• % (+ $"4 ( 6
8.3.2 Q`"44"* y T 6 u" «
$5 ^kuŠ
$"•: + $"4 ( 8.3.15 Ó%u "439*uy —y4/35 « ~ %5 ôŠ
( $"4 (
$"• + By 8.3.37 $6x95 ≍$≍^… S «, *—;"'ki/3 and *u 0y are #"â, but both are negated
6
by (u"þ) ^$"4" ( 2.
9 u´Z5 «, to give -wo1
^“1
$"-"4 (
259
‹i-( *+5 8) w0'5 – 2
6 ( w0' starting from

At the beginning of *+ section, there is a small section of T$
6 «( , until 6.1.76 ô"t"Ò" «.
6.1.71 •€Š *^*T $| *T T$
[*u*q `k ']( 6.1.73 ¶1 S « ~ •€Š T$6 ( *‹T"3"' (

6 ( w0' becomes Qt-Qu3u of a short vowel when followed by ¶ (.
T$
6 7
•€ ¶(
1
6 ( w0'
T$
(Qt-Qu3u)
6 ( 1/1 *‹T"3"' 7/1

¶1 7/1 S 0 « ~ •€Š 6/1 T$ (

• ¶1 7/1 – In ^% â'/.
• 6 «( .
S 0 – This connects with 6.1.71 •€Š *^*T $| *T T$
• 6 «( ; •€ is defined by 1.2.27 ù$"i9›
•€Š 6/1 – From 6.1.71 •€Š *^*T $| *T T$ €o/•‚y 6T5;
in v"431 90" sé/.
• 6 ( 1/1 – This is w0'; from 6.1.71 •€Š *^*T $| *T T$
T$ 6 «( ; ƒ before $ ( is ƒí"%&"‡ y, not pT; ( thus
it does not bring u&ys.
6 ( 1/1 «
[LSK] •€Š 6/1 ¶1 7/1 T$
6 ( is the w0' for •€ when ¶ ( follows.
T$
[LSK] *2u•"3" - The shadow of Śiva.
*2uŠ ¶"3" *2u•"3" «
( ¶"3" + • 6
*2u + ™ +
*2u + ¶"3"
6 ( + ¶"3"
*2u + T$ 6 ( with the help of 1.1.46 w±t… »*$T… «
6.1.73 ¶1 S« ~ •€Š T$
*2uo ( + ¶"3" 8.2.39 ði" —29›t1« ~ ôŠ
( ¶"3"
*2u— + 6 š5«
8.4.40 Æ95 š4" 6
( ¶"3"
*2uS + 8.4.55 Ó*% S « ~ ði" S% (
260
‹¢ *+5
[*u*q `k ']( 6.1.76 ô"t"Ò" « ~ o/•"yT ¶1 6 ( *‹T"3"' (

( T$
6 ( w0' optionally becomes Qt-Qu3u of a long vowel at the end of ô when
T$
followed by ¶ (. 5 7
ô o/• y ¶(
1
Optional
6 ( w0'
T$
(Qt-Qu3u)
( u" 0 « ~ o/•"yT 5/1
ô"t"T 5/1 6 ( 1/1 *‹T"3"' 7/1
( ¶1 7/1 T$ (

• (
ô"t"T 5/1 – ôŠ Qt5 ô"t5 (6T), T¼"T«;( adjective to o/•"yT; ( in ^ku^y Í'/.
• (
o/•"yT 5/1 ( long vowel defined by 1.2.27 ù$"i9›
– From 6.1.75 o/•"yT «; €o/•‚y 6T5« .
• ¶1 7/1 – From 6.1.73; in ^% â'/.
• 6 ( 1/1 – This is w0'; from 6.1.71 •€Š *^*T $| *T T$
T$ 6 «( .
( is
Here, since there is no 6th case-ending word, ^*%n"s" “ƒn3*4oË2 1 ^Í'/*4oË29 Ri/3"4 «”
6 ( is exceptionally considered to be Qu3u of o/•.y This is because such
applicable. However, T$
6
usage is seen in ^"*&*4’s own sūtra 2.4.25 *un"s" 41 " %"•"3"2"i"*42"4"' ( Such instance is
«.
called !"^$, indicator of certain forms.
( ô"t"T 5/1
[LSK] o/•"yT 5/1 6 ( 1/1 u" 0 «
( ¶1 7/1 T$
6 ( is optionally w0' when ¶ ( follows.

After o/0y at the end of ô, T$
[LSK] iy/•"3", iy/¶"3" -
iz"5 ¶"3" iy/•"3", iy/¶"3" (6T)«
( ¶"3" + • 6
iy/+ ™ +
iy/+ ¶"3"
6 ( + ¶"3"
iy/+ T$ 6 ( with the help of 1.1.46 w±t… »*$T… «
6.1.76 ¶1 S« ~ •€Š T$
iy/o ( + ¶"3" 8.2.39 ði" —29›t1« ~ ôŠ
( ¶"3"
iy/— + 6 š5«
8.4.40 Æ95 š4" 6
( ¶"3"
iy/S + 8.4.55 Ó*% S « ~ ði" S% (
261
p*T ‹¢ *+5 -
Thus ends the section of consonant sandhi.
Summary of ‹¢ *+5
( the Q4u|6 *§5 from 8.2.108 to 8.4.68, the end of QU"j"3/.
*‹T"3"' is
6 U•
1) š•, 6 - *+5
6 š5«
8.4.40 Æ95 š4" 6 General rules
8.4.41 U4
6 " U56 «
8.4.42 4 ô"t" 9%4"' «(
8.4.43 T95 *s « 6 and U•
Conditions where š• 6 are prohibited
8.4.44 2"T «(
2) ô"t-—Ÿ•- *+5
8.2.39 ði" —29›t1 « ~ ôŠ It has a wide application.
6
3) Q44"* $- *+
6
8.4.45 3%9›44"* 6 $9 u" «
$1 ›44* Ex. ‘T %6 "*%5, ‘T/%"*%5
6
4) Last section *+5

8.4.60 T9*iy « ~ ^% u&y5 Ex. T 35, *uÒ" • nT1
y « ~ u&y5
8.4.61 ƒo5 v"Æf95 ^kuŠ Ex. ƒô"4', ( ƒ§f4' (
8.4.62 ð39 ‹9›ÕT%Š"' «( ~ ^kuŠ
y u&y5 Ex. u"Ï•*%5, Q €5, Š"dO€5, %"}kT,1 QhÆ/
8.4.63 2Ÿ¶9›*» « ~ ð35 QÕT%Š"' ( Ex. ./'• %"S"3y5
6
5) Q4€"%- *+5
6
8.3.23 '9›4€"%5 « ~ ôŠ Ex. ‹*% u¡1
6
8.3.24 4š"ô"tŠ ð*i « ~ '5 Q4€"%5 Ex. '4"*
8.3.25 '9 %"*— '5 ò… « ~ '5
8.3.26 ‹1 '^%1 u" « ~ '5 '5 6
Qûos to Q4€"%
8.3.27 4^%1 45 «
6
8.4.58 Q4€"%Š ^% u&y5 « 6
Changes on Q4€"%
8.4.59 u" ô"tŠ «
262
‹¢ *+5
6) w0'5
1.1.46 w±t… »*$T… « ^*%n"s" `k ' (
8.3.28 6 ( 2*% «
&95 $6q$
8.3.29 ¸5 * q»6 ( «
8.3.30 4š « ~ * q»6 (
6 ( « ~ 45
8.3.31 *2 T$
(
6 *4e'
8.3.32 ™'9 •€"o*S ™'& «(
7) 2nd 7• •-#$%&' (
6
8.3.2 Q`"44"* y T 6 u" «
$5 ^kuŠ They give two options in this section:
6
8.3.4 Q44"* (
$"T ^%9›4 6
€"%5 « 6
Q44"* 6
$ or Q4€"%
8.3.5 6 « ~ 7•
'5 *»
6 Ó[†%1 « ~ 7•
8.3.6 ^'5
8.3.7 Ÿ¶Z#2"4 «( ~ 7• Q†%1 Conditions where 7• • occurs.
( « ~ 7•
8.3.10 4Q4 ^1
8.3.12 $"4"l1*¸T1 « ~ 7•
8.3.15 Ó%u "439*uy —y4/35 « ~ %5 Change to *u 0y
8.3.34 *u —y4/3Š 5 «
8.3.36 u" 2*% « ~ *u —y4/3Š *u —y4/35 Changes of *u 0y
8.3.37 $6x95 ≍$ ≍^… S « ~ *u —y4/3Š
8) T06 "0'5
6.1.72 *‹T"3"' «(
6 (
6.1.73 ¶1 S « ~ •€Š T$ 6 ( is *4e after •€. Ex. p•*T, 0•*T, 3•*T
T$
6 (
6.1.74 w '"™9š « ~ T$ *4e after these two ôs. Ex. w•"o3*T, '"*•oT (
6 (
6.1.75 o/•"yT «( ~ T$ *4e after Qô"t o/•.y Ex. {1•*T
6 (
6.1.76 ô"t"Ò"« ~ T$ *u$® after ô"t o/•.y Ex. iy/•"3", iy/¶"3"
263
Q‡ *u 0y- *+5
Technically speaking, *u 0y *+ is about *u$"%, modification, of *u 0y, which is done

by 8.3.34 *u —y4/3Š 5« and its following Qû"o sūtras.
However, in the *u 0y *+ section of i•*6 }"t$…'o6 /, sūtras taught are about
modifications of a letter which would have been *u 0y if Qu "4 (absence of letter) was
following. In this type of modifications, the letter never becomes *u 0y.
This sūtra is already seen in the last section. After getting *u 0y by 8.3.15
Ó%u "439*uy —y4/35«, how the *u 0y undergoes *u$"% is told by the following two sūtras.
[*u*q `k ']( 8.3.34 *u —y4/3Š 5 « ~ Ó*% *‹T"3"' (

( the substitute for *u 0y followed by Ó%.(
is
6 7
*u —y4/3 Ó% (
1
(
(
*u —y4/3Š 6/1 5 1/1 « ~ Ó*% 7/1 *‹T"3"' 7/1
is
• Ó*% 7/1 – #e"‹"%5 Ó%,( all the hard consonants; in ^% â'/.
[LSK] Ó*% 7/1 «

( the substitute in the place of *u 0y when Ó% (follows.
is
6
[LSK] *uŽ¥"T" - ViJ u is the protector.
*uŽ56 + `"T"
( `"T"
*uŽ 6 + 8.3.34 *u —y4/3Š 5 «
264
‹¢ *+5
The next sūtra is also already seen in the previous section.
[*u*q `k ']( 8.3.36 u" 2*% « ~ *u —y4/3Š *u —y4/35 *‹T"3"' (

*u 0y is optionally the substitute for *u 0y when followed by 2%.( The other option is . (
6 7
*u —y4/3 2% (
1
*u —y4/3
Optional
(
u" 0 2*% 7/1 « ~ *u —y4/3Š 6/1 *u —y4/35 1/1 *‹T"3"' 7/1
• (
2*% 7/1 – #e"‹"%5 2%,( sibilants 2, ( s and ; ( in ^% â'/.
• *u —y4/35 1/1 – From 8.3.35 2^ y%1 *u —y4/35«; this is wo12; this is brought by the S to
eliminate the possibility for ( 2 by 8.3.34 *u —y4/3Š 5«.
wo1
[LSK] 2*% 7/1 *u 0yŠ 6/1 *u 0y5 1/1 u" 0«

*u 0y is optionally the substitute in the place of *u 0y when 2% (follows. The other
option is . (
[LSK] ‹*%5 21T,1 ‹*%Ÿ21T 1 - Hari sleeps.
‹*%5 + 21T 1
‹*%5 + 21T 1 8.3.36 u" 2*% « ~ *u —y4/3Š *u —y4/35
^“1
‹*% ( 21T 1
+ 8.3.34 *u —y4/3Š 5 «
( 21T 1
‹*% 2 + 6
8.4.40 Æ95 š4" 56 «
265
[*u*q `k ']( 8.2.66 6 7• : « ~ ôŠ

—s9
ô"t is (
( replaced by 7• . The last letter of —s6 2² is also replaced by 7• .
6
ô (
1
7•
6 6/1 7• : 1/1 « ~ ôŠ 6/1

—s95
• 6 6/2 – 5 S —k5 S
—s95 6 (ID), T395; Q of
—s… (
is ƒí"%&"‡ y; —s6 means friend; adjective
to ôŠ.
• 7• : 1/1 – This is wo12; ƒ• is pT. (
• ôŠ 6/1 – From 8.1.16 ôŠ«; in v"431 90" sé/.
Since 6 is adjective to ôŠ, Tot*u*q by 1.1.72 314 *u*qÆotŠ« is applied. They

—s95
(
are understood as “ $"%"tŠ ôŠ” and “ —s6 -2²"tŠ ôŠ”, then 1.1.52 Qi9›•Š« is
applied.
[LSK] ô"tŠ 6/1 Š 6/1 —s5 (

6 6/1 S 0 7• : 1/1 Š"T III/1 -
7• is optionally the substitute in the place of ô"t (
and (
the last letter of —s6 2².
Example is given in the next sūtra.
266
‹¢ *+5
[*u*q `k ']( 6.1.113 QT9 %9%‚T"o‚

6 (
T16 « ~ ƒT Q*T
7• is replaced by short ƒ, when 7• is between two short Qs.
5 6 7
QT ( 7• QT (
1
ƒ
( Q‚ 6T1 7/1 « ~ ƒT 1/1

QT5 5/1 %95 6/1 Q‚ 6T"T 5/1 ( Q*T 7/1

• QT5 5/1 – #"*T^*o$ is QT; ( T^%$%& is for TB"i-*43' by 1.1.70 T^%ÆB"iŠ«; in ^ku^y Í'/.
• %95 6/1 – #"*T^*o$ is 7• ; this excludes %1Ý without ƒ• pT; ( in v"431 90" sé/.
• (
Q‚ 6T"T 5/1 – 4 ‚ 6T5 Q‚ 6T5 (NT), T¼"T«,( adjective to QT5 .
• Q‚ 6T1 7/1 – 4 ‚ 6T5 Q‚ 6T5 (NT), T*¼4«,( adjective to Q*T.
• (
ƒT 1/1 – This is wo12; from 6.1.111 xT ƒT«;( T^%$%& is just for clarity, as it is *uq/3'"4,
thus does not indicate its u&y.
• Q*T 7/1 – From 6.1.109 ‘™5 ô"t"o*T «; #"*T^*o$ is QT; ( T^%$%& is for TB"i-*43' by 1.1.70
T^%ÆB"iŠ«; in ^% â'/.
7• is told in the section of 7• • starting from 8.2.66 6 75, which is in *`^"o/. In this
—s9
(
sūtra, 7• • is not Q* }uT because of "'Wy• of the word %95.
(
‚ 6T is also told in *`^"*o, in the section starting from 8.2.82. Q‚ 6T"T and Q‚ 6T1 in this
sūtra also indicate that they are not Q* }uT. (
( QT5 5/1 ^%Š 6/1 %95 6/1 ƒ5 1/1 Š"T III/1

[LSK] Q‚ 6T"T 5/1 ( Q‚ 6T1 7/1 Q*T 7/1 «
ƒ is the substitute in the place of 7• which is preceded by a short Q, when a short Q
follows.
267
[LSK] *2u9›Îy5-
*2u + • 6 + QÎy5 (
^6 -#e3 (
• 6 is suffixed by 4.1.2 €…— '…S ...«
( QÎy5
*2u + ( 1.3.2 ƒô12›1 —44"*
ƒ• is pT by 6 $ pT«,( Q4R+i9^5
6 by 1.3.9 TŠ i9^5«
6
Being Rt, 6
ô !" is given by 1.4.14 *â™t ô' «(
*2u7• + QÎy5 8.2.66 6 7• : « ~ ôŠ with the help of 1.1.52 Qi9›•Š «
—s9
*2u% (+ QÎy5 ( 1.3.2 ƒô12›1 —44"*
6 by 1.3.9 TŠ i9^5«
*2u ƒ + QÎy5 (
6.1.113 QT9 %9%‚ 6T"o‚ 6T1 « ~ ƒT Q*T
*2u9 + QÎy5 6 « ~ Q*S *‹T"3"' ^k( u^y %395 ‘$5
6.1.87 w &5
*2u9›Îy5 6.1.109 ‘™5 ô"t"o*T « ~ ^ku5y *‹T"3"' ^k( u^y %395 ‘$5
Note that Q + ƒ inside ô has precedence to ƒ + Q between two ôs, because $"3y
inside ô is Qt%¹, more intimate, than $"3y between two ôs, which is called R*‹%¹.
268
‹¢ *+5
[*u*q `k ']( (
6.1.114 ‹*2 S « ~ Q‚ 6T"T QT5 %95 ƒT (
The replacement of 7• with ƒ happens also when what is following is ‹2. (
5 6 7
QT ( 7• ‹2 (
1
ƒ
( %95 6/1 ƒT 1/1

‹*2 7/1 S 0 « ~ QT5 5/1 Q‚ 6T"T 5/1 (

• ‹*2 7/1 – #e"‹"% ‹2; ( all the soft consonants; in ^% â'/.
• S 0 – This connects with the previous sūtra, as told by “T‡"” in u|*§.
• QT5 5/1 – #"*T^*o$ is QT; ( T^%$%& is for TB"i-*43' by 1.1.70 T^%ÆB"iŠ«; in ^ku^y Í'/.
• (
Q‚ 6T"T 5/1 – 4 ‚ 6T5 Q‚ 6T5 (NT), T¼"T«,( adjective to QT5 .
• (
ƒT 1/1 – This is wo12; from 6.1.111 xT ƒT«.(
[LSK] T‡" 0 «
(
ƒ is the substitute in the place of 7• which is preceded by a short Q, when ‹2 follows.
[LSK] *2u9 u|5-

*2u + • 6 + u|5 (
^6 -#e3 (
• 6 is suffixed by 4.1.2 €…— '…S ...«
( u|5
*2u + ( 1.3.2 ƒô12›1 —44"*
6 by 1.3.9 TŠ i9^5«
6
Being Rt, 6
ô !" is given by 1.4.14 *â™t ô' «(
*2u7• + u|5 8.2.66 6 7• : « ~ ôŠ with the help of 1.1.52 Qi9›•Š «
—s9
*2u% (+ u|5 ( 1.3.2 ƒô12›1 —44"*
6 by 1.3.9 TŠ i9^5«
*2u ƒ + u|5 (
6.1.113 QT9 %9%‚ 6T"o‚ 6T1 « ~ ƒT Q*T
*2u9 + u|5 6 « ~ Q*S *‹T"3"' ^k( u^y %395 ‘$5
6.1.87 w &5
269
[*u*q `k ']( y 39›*2 « ~ %95

8.3.17 n9n09Q•9Q^kuŠ
7• which is preceded by n9, n09, Q•9, or Qu&y is replaced by 3, ( when Q2 is
( following.
6 7
n9/n09/Q•9/Qu&y 7• Q2 (
1
3(
y 6/1 35 1/1 Q*2 7/1 « ~ %95 6/1

n9-n09-Q•9-Q^kuŠ
• y 6/1 – n95 S n095 S Q•95 S Q5 S n9n09Q•9w5 (ID)«; n9 , ( n09 , ( and Q•9 (
n9-n09-Q•9-Q^kuŠ
are exclamations ending with ; ( here in this sūtra, the forms without , ( namely, n9, n09,
Q•9, are compounded; *+-Qn"u between n09 and Q•9, as well as Q•9 and Q is …`,
belonging to sūtra; T1 ^kuË 3¼"T ( 5 n9-n09-Q•9-Q^ku5y (115B), TŠ«; this can be understood
y , n09^kuŠ
as n9^kuŠ y Q•9^kuŠ
y , and Qu&yûk Š
y ; adjective to %95.
• 35 1/1 – This is wo12; Q of 3 is ƒí"%&"‡ y.
• Q*2 7/1 – #e"‹"% Q2; ( all the vowels and soft consonants; in ^% â'/.
• S 0 – This connects with the previous sūtra, as told “T‡"” in u|*§.
( uŠ
[LSK] ‘TT-^k y 6/1 %95 6/1 3-wo125 1/1 Q*2 7/1 «
( the substitute in the place of 7• which is preceded by n9, n09, Q•9, or Qu&y, when
3 is
(
Q2 follows.
(
[LSK] n9 n09 (
Q•9 ( 0 -Qt"5 1/3 *4^"T"5 1/3«
p*T
(
These exclamations are -ending *4^"Ts, listed in S"*o0&.
( used for addressing several persons.
n9 is
( used for addressing Bhagavān.
n09 is
( used for addressing ^"*^4s.(
Q•9 is
270
‹¢ *+5
n9 , ( n09 , ( Q•9 are

( in S"*o0&. *4^"T- !" is given by 1.4.57 S"o39› «1 .
(
To *4^"T !$ words, QZ3- !" is given by 1.1.37 €%"*o*4^"T'Z3' «.
They all having meaning, #"*T^*o$- !" is given by 1.2.45 Q‡uy oq"T%6 #e35 #"*T^*o$' «(
(
By being #"*T^*o$, ^6 -#e3 ( ~ #"*T^*o$"T (
is added by 4.1.2 €…— '…S …«
(
By being QZ3 !$ word, ^6 -#e3 is elided by 2.4.82 QZ3"o"`^6 5«
6
By being Rt, 6
ô- !" is given by 1.4.14 *â™t ô' «(
By getting ô !", 8.2.66 6 7• : « is applicable.
—s9
( o1u"5
n9 +
n97• + o1u"5 8.2.66 6 7• : « ~ ôŠ with the help of 1.1.52 Qi9›•Š «
—s9
( o1u"5
n93 + y 39›*2 «
8.3.17 n9n09Qn9Q^kuŠ
( %95 6/1 3•1 7/1 $| T1 7/1 -

[LSK] T1s"' 6/3
Of these three *4^"Ts (n9, n09, and Qn9), 7• is replaced by 3 –( (to be continued)
i•*6 }"t$…'o6 /$"% shows examples for those three *4^"Ts first.
271
[*u*q `k ']( 8.3.22 ‹*i uËs"' «( ~ n9n09Q•9Q^kuŠ

y 3Š i9^5
(
3 which is preceded by n9, n09, Q•9, or Qu&y is elided, when ‹i ( is following,
according to the opinion of all the teachers.
6 7
n9/n09/Q•9/Qu&y 3( ‹i (
1
i9^
( « ~ n9-n09-Q•9-Q^kuŠ
‹*i 7/1 uËs"' 6/3 y 6/1 3Š 6/1 i9^5 1/1
• ‹*i 7/1 – #e"‹"% ‹i;( all the consonants; in ^% â'/.
• (
uËs"' 6/3 ( «; in the opinion of all the grammarians; this makes the rule
– wS"3"y&"' 'T1
*4e.
• y 6/1 – Same as above; in v"431 90" sé/.
n9-n09-Q•9-Q^kuŠ
• 6
3Š 6/1 – From 8.3.18 Z9iy•#3ÉT%5 ( 6th case is taken; Q of 3 is
2"$»"34Š«, only 3 in
ƒí"%&"‡ y; in v"431 90" sé/.
• i9^5 1/1 – This is wo12; from 8.3.19 i9^5 2"$ÞŠ«.
[LSK] n9-n09-Q•9-Q^kuŠ (
y 6/1 3Š 6/1 i9^5 1/1 Š"T III/1 ‹*i 7/1 «
(
i9^ is the substitute in the place of 3 which is preceded by n9, n09, Q•9, or Qu&y, when
‹i ( follows.
[LSK] n9 o1u"5 « n09 4'Æ1 « Q•9 3"*‹-

( o1u"5
n9 +
n97• + o1u"5 8.2.66 6 7• : « ~ ôŠ with the help of 1.1.52 Qi9›•Š «
—s9
( o1u"5
n93 + y 39›*2 «
8.3.17 n9n09Qn9Q^kuŠ
n9 + o1u"5 8.3.22 ‹*i uËs"' «( ~ n9n09Qn9Q^kuŠ
y 3Š i9^5
272
‹¢ *+5
[*u*q `k ']( 6 « ~ Q‹4 (

8.2.69 %9› *^
( replaced by % (when ^6 is
The last letter of Q‹4 is ( not following.
6 7
of Q‹4 ( 4( not ^6 (
1
%(
%5 1/1 Q *^ (
6 7/1 « ~ Q‹4 6/1

• %5 1/1 – This is wo12; Q of % is ƒí"%&"‡ y.
• ( ^6 (NT),
6 7/1 – 4 ^6 Q
Q *^ ( T*¼4«;( in ^% â'/.
• (
Q‹4 6/1 – From 8.2.68 Q‹4«;( in v"431 90" sé/; *un*´ is â, elided.
[LSK] Qo5 6/1 %1Ý-wo125 1/1 4 0 T 6 0 *^

6 7/1 «
% (is the substitute in the place of the last letter of Q‹4, ( when ^6 is
( not following.
y6 5-
[LSK] Q‹%‹5 « Q‹0&
( • 6 + Q‹4 +
Q‹4 + ( • 6 *Ò7*´5 by 8.1.4 *4eu/`395 « ~ uyŠ Ò1
( Q‹4 (
Q‹4 + 7.1.23 €'94 y^ 6 $"T «( ~ $ (
A #e3 which is elided by $-( elision does not become the *4*'§ for a rule which
causes change on its Q¹ by 1.1.163 4 'T"¹Š «
Q‹% (+ Q‹4 ( (
6 « ~ Q‹4 with
8.2.69 %9› *^ the help of 1.1.52 Qi9›•Š «
Q‹% (+ Q‹% ( (
6 « ~ Q‹4 with
8.2.69 %9› *^ the help of 1.1.52 Qi9›•Š «
Q‹% (+ Q‹5 8.3.15 Ó%u "439*uy —y4/35 « ~ ôŠ
273
[*u*q `k ']( 8.3.14 %9 *% « ~ i9^5

% (is elided when another % (follows.
6 7
%( %(
1
i9^
%5 6/1 *% 7/1 « ~ i9^5 1/1

• %5 6/1 – In v"431 90" sé/;
• *% 7/1 – In ^% â'/.
• i9^5 1/1 – This is wo12.
[LSK] %1ÝŠ 6/1 %1Ý1 7/1 ^%1 7/1 i9^5 1/1 «

i9^ is substitute in the place of % (when % (follows.
274
‹¢ *+5
[*u*q `k ']( y o/•Ì›&5 «

6.3.111 Oi9^1 ^kuŠ
(
When ñ ( or % (is elided, the Q& before the cause of that elision gets elongated.
6 7
Q& ( Oi9^
1
o/• y
Oi9^1 7/1 ^kuŠ

y 6/1 o/•5y 1/1 Q&5 6/1 «
4 words in the `k , no word is required as Q4u|6 *§.
• (
Oi9^1 7/1 – ñ ( S % S O (ID)«; Q of % (is ƒí"%&"‡ y; O… 2/2 i9^3*T Oi9^5 (UT)«; O + ç» ( + ^ +
( Q&«;( indicating that which caused the elision of ñ ( or %;( Oi9^5 can be said to be
*&S +
Oi9^*4*'§5, a cause for the elision of ñ ( or %;( ñ ( and % (are elided by 8.3.13 ñ9 ñ1 i9^5« 8.3.14
%9 *%«, thus Oi9^*4*'§5 can be either ñ ( or %;( in ^% â'/.
• y 6/1 – Before the Oi9^; adjective to Q&5.26
^kuŠ
• Q&5 6/1 – #e"‹"% Q&; ( being Q*uq/3'"4, by 1.1.69 Q&*6 o„u&yŠ S"#e35«, Qu&y pu&y and ƒu&y
are indicated; in v"431 90" sé/;
• o/•5y 1/1 – This is wo12. ^*%n"s" 1.1.50 v"4›1 t%T'5« is required.
( ) ^kuŠ
[LSK] ñ-%1Ý395 6/2 i9^*4*'§395 7/2 ( T95 7/2 T"8"' 5/2 (
y 6/1 Q&5 6/1 o/•5y 1/1 Š"T III/1 «
o/•5y is substitute in the place of Q&, ( which is before the cause of elision of ñ ( or %.(
26
y «, by hearing Oi9^1 in 7th case, ^kuŠ
Because of 1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ y is understood. However,
the mentioning of ^kuŠ 6

y is redundant because it prevents ƒ§%ô1 from 6.3.1 Q 0§%ô1
« to come. Thus,
Oi9^ does not have to be in the ƒ§%ô.
275
(
Since Q& consists of three types of letters, examples are also of three types: Qu&y, pu&y, and
ƒu&y followed by ñ ( or % (which is the cause of elision of ñ ( or %.(
6 %'T1 « Again, he is delighted.
[LSK] ^4"
6 (+ %'T1
^4% 6 (is %1Ý"t-QZ3 word.
^4%
^46 + %'T1 8.3.14 %9 *% « ~ i9^5
6 + %'T1
^4" y o/•Ì›&5«, with the help of 1.1.50 v"4›1 t%T'5«
6.3.111 Oi9^1 ^kuŠ
[LSK] ‹%/ 1/1 %Å5 1/1 « Hari is beautiful.

( %Å5
‹*% + After #"*T^*o$ ‹*%, • 6 is suffixed. ô !" is gained by 1.4.14 *â™t
6 (
ô' «.
‹*%7• + %Å5 8.2.66 6 75« ~ ôŠ
—s9
‹*%% (+ %Å5 6
Q4R+i9^5 «
‹*%+ %Å5 8.3.14 %9 *% « ~ i9^5
‹%/ + %Å5 y o/•Ì›&5«, with the help of 1.1.50 v"4›1 t%T'5«
6.3.111 Oi9^1 ^kuŠ
[LSK] 2fk 1/1 %"—T1 1/1 « Śambhu shines.
Exactly in the same manner as above.
ô$| e is done for “Q&5”.

( ? T|ñ5 1/1 « u|ñ5 1/1 -
[LSK] Q&5 6/1 *$' 1/1
( ´
T|U• *‹ "3"' + 3.2.102 *4é" « ~ nkT 1 #e3 ^%š q"T95
T|‹ ( + T 6
Q4R+i9^5
T|ñ ( + T 8.2.31 ‹9 ñ5 « ~ ð*i
T|ñ ( + q 8.2.40 ðsÆ‡9qÌ›q5 «
T|ñ ( + ñ 8.4.41 U4
6 " U56 « ~ Æ95
T| + ñ 8.3.13 ñ9 ñ1 i9^5 «
( not Q&, ( hence 6.3.111 Oi9^1 ^kuŠ
x after T is y o/•Ì›&5« is not applied.
In the same manner, this sūtra does not apply to u|ñ5, derived from u|U• ƒ±'4 1 to raise,
with ´.
276
‹¢ *+5
Now, i•*6 }"t$…'o6 / gives a set-up with sūtras covered in this section, in order to
introduce the next sūtra.
(
[LSK] “'4 %‡5” p*T 0 Q` 0 7• •1 7/1 $| T1 7/1 “‹*2 S (6.1.114)” p*T 0 ƒ•1 7/1 “%9 *% (8.3.14)” p*T 0 i9^1 7/1
S 0 #"â1 7/1 «
(
In the case of “'4 %‡5”, after making 7• • (by 8.2.66 6 7• :«), there are two
—s9
possibilities: ƒ• for 7• by 6.1.114 ‹*2 S « or i9^ for %1Ý by 8.3.14 %9 *% «.
'4 5 %‡5 '49%‡5 (6T) «
( ™ +
'4 + ( %‡ + • 6 2.2.8 sé/ « ~ '" 5 T°7s5,
6 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
( %‡
'4 + 6 q"T#6 *T^*o$395 « ~ $ (
2.4.71 ^9
'47• + %‡ 8.2.66 6 7• :« ~ ôŠ
—s9
'4 ƒ + %‡ (
ƒ• for 7• by 6.1.114 ‹*2 S « ~ Q‚ 6T"T QT5 %95 ƒT (
OR
'4 + %‡ i9^ for %1Ý by 8.3.14 %9 *% « ~ i9^5
Since there are two sūtras which are applicable in one place at the same time, we
need a ^*%n"s" `k . The next sūtra gives a criterion for judgment.
[^*%n"s" `k ']( 1.4.2 *u#*Ts1q 1 ^% $"3y' «(

When two sūtras are applicable in one place at the same time, the one with bigger
number is taken.
( $"3y' 1/1
*u#*Ts1q 1 7/1 ^%' 1/1 ( «

• *u#*Ts1q 1 7/1 – When two sūtras are applicable in one place; in *T â'/.
• ^%' (1/1 – The latter one; adjective to $"3y'. (
• $"3y' ( 1/1 – The effect.
( $"3y' 1/1
6 -Ri-*u%9q1 7/1 ^%' 1/1
[LSK] TÞ ( Š"T III/1
( «
When two sūtras are applicable in one place at the same time, the change which has
a latter number should be applied.
277
6 Ri*u%9q, is explained in *»å&/.

The word in the sūtra *u#*Ts1q is told as TÞ
[LSK *»å&/ !] QÕ` 0 QÕ` 0 i -Qu$"2395 6/2 ( `k 395 6/2) ‘$` 0 306 ^T 0( #"*â5 1/1 TÞ
6 Ri*u%9q5 1/1«
Applicability in one place at the same time of two sūtras, which have their own
6 Ri*u%9q5, contradiction or quarrel of two things which
sphere of operation is called TÞ
have the same strength.
6 %'T1, only 8.3.14 %9 *% « is applicable and 6.1.114 ‹*2 S « has no scope.
In the case of ^4"
Similarly, in the case of *2u9 u|5, only 6.1.114 ‹*2 S « is applicable and 8.3.14 %9 *% «has no
scope. This means that they have their own scope of operation, Qu$"2.
[LSK] p*T 0 i9^1 7/1 #"â1 7/ 1 «

By this sūtra 1.4.2 *u#*Ts1q 1 ^% $"3y'«,( because 8.3.14 %9 *% « ~ i9^5 has bigger number
than 6.1.114 ‹*2 S « ~ ƒT, ( i9^ is applicable, but…
(
[LSK] “^ku`y "* }' (8.2.1)” ( ƒ•' 1/1
p*T 0 “%9 *% (8.3.14)” p*T 0 QŠ 6/1 Q* }•"T 5/1 ( ‘u 0 «
( the sūtra 8.3.14 %9 *% « is as good as non-existent. Thus
Because of 8.2.1 ^ku`y "* }' «,
ƒ• by 6.1.114 ‹*2 S « is to be taken.
[LSK] '49%‡5 -
( %‡
'4 +
'47• + %‡ 8.2.66 6 7• :« ~ ôŠ
—s9
By 1.4.2 *u#*Ts1q 1 ^% $"3y'«,( 8.3.14 %9 *% « was #"â. However, from the standpoint of
(
6.1.114 ‹*2 S «, 8.3.14 %9 *% « does not exist by the arrangement taught by 8.2.1 ^ku`y "* }' «.
Thus ƒ• by 6.1.114 ‹*2 S « takes place.
'4 ƒ + %‡ (
6.1.114 ‹*2 S « ~ Q‚ 6T"T QT5 %95 ƒT (
'49+ %‡ 6 « ~ ‘$5 ^ku^y %395 Q*S *‹T"3"' (
6.1.87 w &5
'49%‡
More examples for this ^*%n"s" 1.4.2 *u#*Ts1q 1 ^% $"3y' «( will be seen in later sections.
278
‹¢ *+5
Note:
When two sūtras are applicable in one place, the following ^*%n"s" explains which
sūtra is stronger:
6
(^*%n"s") ^ku^y %-*4e-Qt%¹-Qû"o"4"'§%9§% Ri/35«
^% `k to ^ku y `k , *4e to Q*4e, Qt%¹ to R*‹%¹, and Qû"o to ƒ„0y. Among these four
types of comparison, the latter is the more effective.
Four types of comparison:

1) ^ku-y ^%
This is told by 1.4.2 *u#*Ts1q 1 ^% $"3y' «(
2) *4e-Q*4e
$| T"$| T-# ¹/ 35 *u*q5 5 *4e5« A *u*q which is applicable both before and after
another *u*q is applied.
Let both *u*q A and *u*q B be applicable in one place at the same time. If *u*q B is
applicable even after applying *u*q A, but *u*q A becomes non-applicable after
applying *u*q B, then *u*q B is considered to be *4e and *u*q A is considered to be Q*4e.
3) Qt%¹-R*‹%¹
*u*q between q"T 6 and *T™-( #e3 is Qt%¹, while *u*q between ƒ^ 0y and q"T 6 is R*‹%¹.
*u*q between ƒ^ 0y and q"T 6 is Qt%¹, while *u*q between another ô and ƒ^ 0y is R*‹%¹.
4) ƒ„0y-Qû"o
When *u*q A has no applicability outside the scope of *u*q B, *u*q A is said to be
*4%u$"2 or Qû"o, and has to be taken, while *u*q B is called ƒ„0y.
279
[*u*q `k ']( 6
6.1.132 ‘T§o95 i9^9›$9%4È ( '" 1 ‹*i «
• 6 of ‘s5 and 5 is elided when ‹i ( follows. They should not be ‘s$5, $5, Q4s1 5, or
Q 5.
(
TT/‘TT ( 6 7
not with $, •6 ‹i (
not in 4Ø '"
1
i9^
‘T§o95 6/2 i9^5

6 1/1
Q$95 6/2 Q4Ø '" 1 7/1 ‹*i 7/1 «
• ( TT S
‘T§o95 6/2 – ‘TT S ( ‘T§o… (ID), T395«; ¬+1 sé/ to 6 of i9^5.
6
• 6
i9^5 6
1/1 – This is wo12; 95 i9^5 i9^5 (6T)«.27
• Q$95 6/2 – Q*u±'"45 $ ( 3395 T… Q$… (NB), T395«; adjective to ‘T§o95.
• Q4Ø '" 1 7/1 – 4 4Ø '" 5 Q4Ø '" 5 (NT), T*¼4«;( in Q*q$%&1 â'/.
• ‹*i 7/1 – #e"‹"%5 ‹i;( all the consonants; in ^% â'/.
[LSK] Q$$"%395 6/2 ‘T-To95 6/2 35 1/1 :• 6 1/1 TŠ 6/1 i9^5 1/1 ‹*i 7/1 4 0 T 6 0 4Ø '" 1 7/1 «
(
i9^ is the substitute in the place of • 6 of ‘TT and (
TT word, which do not have Q$S-(
#e3 and not in 4Ø '" , when ‹i ( follows.
(
When ‘TT and (
TT are to be suffixed by ,• 6 they are naturally in #‡'" ‘$uS4'. ( Thus
(
their forms are ‘s5 and 5. When ‘TT and ( 4^ 6 $*i¹ and ¥/*i¹ are in #‡'" ‘$uS4', ( • 6 is
TT in
elided by 7.1.23 €'94 y^ 6 $"T«( and 6.1.68 ‹i ( õ"I9 o/•"yT ( *TŠ^|
6 ´ ‹i ( «, respectively. Thus
there is no occasion for this sūtra.
27
6
‘T§o95 is only connected only to one part of i9^5. This kind of connection is called ‘$o12-Qµ35,
and does not fulfil one of the requirements, "'Wy, for qualifying word to be compounded. This can
6
be exempted by being …`, or by dividing the word i9^ into 6 with âsé/ and i9^5.
280
‹¢ *+5
(
Examples for ‘TT and TT, ( respectively.
[LSK] ‘s *uŽ56 «
(
‘s *uŽ 56
‘s *uŽ56 6
6.1.132 ‘T§o95 i9^9›$9%4Ø '" 1 ‹*i «
[LSK] 2f56 «
(
2f56
2f56 6
6.1.132 ‘T§o95 i9^9›$9%4Ø '" 1 ‹*i «
ô$| es for three words in the sūtra are done.

[LSK] Q$95 *$'?( ‘s$9 7L5 «
( €"‡Ë. This Q$S-#e3
There is a T*}T-#e3 called Q$S in ( comes before *» of Q¹. When
(
this Q$S-#e3 (
is suffixed to ‘TT and TT, ( this sūtra does not apply.
( Qo (
( Q$S +
‘T + ( ( »1 5« ~ T*}T"5 #e35 ^%š õ"T"*T^*o$"T (
5.3.71 QZ3 uy4"Ä"'$S #"$
( Q$ + Qo (
‘T + 1.3.3 ‹i•' «( 1.3.9 TŠ i9^5«
‘T$o ( 6 « ~ ^%–^' (
6.1.97 QT9 0&1
Being T*}T"t word, #"*T^*o$- !" is given by 1.2.46 $| §*}T '" "š « ~ #"*T^*o$' (
‘T$o ( + • 6 + 7L5
‘T$ Q + ( 7L5
+ 7.2.102 eo"o/4"'5 « ~ *un´…
‘T$ + ( 7L5
+ 6 « ~ ^%–^' (
6.1.97 QT9 0&1
‘ $+ ( 7L5
+ 7.2.106 To95 5 "u4•395 «
‘s$ + ( 7L5
+ 8.3.59 wo12#e3395 «
6
i9^ does not happen because of the word Q$95.
‘s$ + 7• + 7L5 8.2.66 6 7• :« ~ ôŠ
—s9
‘s$ + ƒ + 7L5 (
6.1.114 ‹*2 S « ~ Q‚ 6T"T QT5 %95 ƒT (
‘s$9 + 7L5 6.1.87 w &5
6 «
( added to To (, it would be: $9 7L5«.

In the same manner, when Q$S is
281
[LSK] Q4Ø '" 1 *$'?( Q 5 *2u5 «28

4 5Q 5«
( To ( + 6
4È +
( To (
4È + 6 q"T#6 "*T$*o$395 « ~ $ (
2.4.71 ^9
Q + To ( 6.3.73 4i9^9 4È5 «
QTo ( + 6 + *2u5
QT Q + ( + *2u5 7.2.102 eo"o/4"'5 « ~ *un´…
QT + ( *2u5
+ 6 « ~ ^%–^' (
6.1.97 QT9 0&1
Q + ( *2u5
+ 7.2.106 To95 5 "u4•395 «
6
i9^ does not happen because of the word Q4Ø '" .1
Q + 7• + *2u5 8.2.66 6 7• :« ~ ôŠ
—s9
Q 5 + *2u5 8.3.15 Ó%u "439*uy —y4/35« ~ ôŠ
In the same manner, Q4s1 5 *2u5«.
[LSK] ‹*i *$'?( ‘s9›` -

Consonant should follow.
When vowel follows, if it is a short Q, 7• • by 8.2.66 6 7• :« and ƒ• by 6.1.113 QT9
—s9
%9%‚ 6T"o‚ 6T1« will take place. If a vowel other than short Q follows, 7• • by 8.2.66 6 7• :« and
—s9
( 2 by 8.3.17 n9n09Q•9Q^kuŠ
3-wo1 (
y 39›*2 «, and optional 3-i9^ by 8.3.19 i9^5 2"$ÞŠ«, then ‘s
25«.
28
Q4Ø '" 1 is # ·#*Ts1q. If it were taken as ^3o6y " , there would be ”Ïœ‹&, i.e. something in the
same —"*T, class, to 4Ø '" should be understood. Then ‘To ( and To ( should be in any '" other
than 4Ø '" . If that were so, ‘s 7L5, or *2u5 would not be possible. This is why Q4Ø '" 1 is
understood as # ·#*Ts1q.
282
‹¢ *+5
[*u*q `k ']( 6.1.134 9›*S i9^1 S1°"o^k%&' «( ~ i9^5

6
6 of “ 5”, which is followed by QS, ( is elided if the measure of a verse is fulfilled
when the elision is done.
5 6/1 Q*S 7/1 i9^1 7/1 S1T 0( ^"o^k%&' 1/1

( « ~ i9^5
6 1/1

• 6
5 6/1 – This is Q4$%& (mimicking of sound) of To ( in 1/1; âsé/-Qt, a word which
ends with an elided 6th case suffix; in ¬+1 sé/ to 6 of i9^5.
6
• Q*S 7/1 – #e"‹"%5 QS; ( all the vowels; in ^% â'/.

• i9^1 7/1 – In *T â'/.
• S1T (0 – “If”.
• ^"o^k%&' ( 1/1 – ^"o"4" ^k%&' (6T)«;
( fulfilment of measurement of verse.
• 6
i9^5 6
1/1 – From 6.1.132 ‘T§o95 i9^9›$9%4Ø '" 1 ‹*i «; this is wo12.
(
[LSK] “ 5” p*T 0 QŠ 6/1 95 6/1 i9^5 1/1 Š"T III/1 Q*S 7/1 ^"o5 1/1 S1T 0( i9^1 7/1 *T 7/1 ‘u 0 ^k[T
Ë III/1 «
i9^ is the substitute of • 6 of “ 5” followed by QS, ( if the measurement of the verse
would fulfilled only when the i9^ is there.
[LSK] '1 "'*u*} #n|*T' «(

This mantra is from xÏu1o 2.24, in —0T/ ¶¡5. The wording without sandhi “ 5 p'"' (
Q*u¸ (*ñ #n|*T'”( … can fit to this measure only if this sūtra 6.1.134 9›*S i9^1 S1°"o^k%&' «( ~
6
i9^5 6
is applied. After i9^, 6 *+ is applied.
0&
[LSK] •s o"2%‡/ %"'5 -
(
This is in Q4U6 ^6 -¶¡5. ( u|*} *+
“ 5 ‘s5 o"2%*‡5 %"'5” with 6.1.134 9›*S i9^1 S1°"o^k%&' «,
y o/•Ì›&5« are applied for o"2%*‡5.
is applied. Notice that 8.3.14 %9 *%« and 6.3.11 Oi9^1 ^kuŠ
[LSK] p*T *u 0y *+5 -

Thus ends the section of *u 0y *+.
283
Summary of *u 0y *+
The topic of *u 0y *+ is well summarized by the *u 0y *+ chart in Sanskrit Grammar

for Vedanta Students Volume 2.
In fact, *u 0y *+ is actually about the change of %1Ý. The causes of the change are:
1. The preceding letter, which can be only vowels.
2. The following letter, which can be any letter.
3. Whether the %1Ý is from 7• or otherwise.
In the chart, the preceding letters are in rows. The following letters are in columns.
In the beginning stage, the whole area of the chart is covered by %1Ý, either of 7• • (a.)
or by itself, such as %1Ý-ending QZ3 words (b.).
a. 7• • from ô"t- $"%5
*2u + • 6 (1/1)
*2u ( 6
1.3.2 ƒô12›1 —44"* ( 1.3.9 TŠ i9^5«
$ pT «,
*2u7• 8.2.66 6 75 «
—s9
*2u% ( 6
1.3.2 ƒô12›1 —44"* ( 1.3.9 TŠ i9^5«
$ pT «,
( itself
b. % by
6 (+ • 6 (1/1)
^4%
6 (
^4% QZ3"o"`^6 5«
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s / ((
w % ((either % (or 7• )
pS (
284
‹¢ *+5
A. For both % (and 7• , when Ó% (follows:
Layer 1: Change of % (to *u 0y

by 8.3.15 Ó%u "439*uy —y4/35«
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q
w % ((either % (or 7•*u) —y4/35
pS (
Note that when Qu "4 (absence of letter) follows, the operation stops here, as in *2u5.
(
Layer 2: Change of *u 0y to /*u 0y/≍
As ƒ„0y, 8.3.34 *u —y4/3Š 5« covers the whole area to change *u 0y to . (
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q
w % ((either % (or 7• ) (
pS (
Then comes its Qû"os:

8.3.36 u" 2*%« gives option for the change of *u 0y to , ( when 2, ( s, ( or (
follows.
8.3.37 $6x95 ≍$ ≍^… S « gives optional *—;"'ki/3 and ƒ^:"4/3 when $6 and ^ 6 follow.
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q (
op. op.
w % ((either % (or 7• )
≍ op. *u —y4/35 ≍
pS (
E.g., Ê5Ó', ( ^45

6 ^45,
6 2"*t5 2"*t5, ^35 .6
285
Layer 3: Change of to ( (
( 2/s
6 and U•
š• 6 š5«
6 by 8.4.40 Æ95 š4" 6 and 8.4.41 U4
6 " U56 « when ( followed by Su0y and »u0y.
is
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q (
op. op.
w 2 7•( ) s (
% ((either % (or
≍ op. *u —y4/35 ≍
pS (
E.g., %"'š, %"'U/$T1, 2"*tŸ2"*t5, %"'aé5.

The rest remains as . ( E.g., %"'Æ%*T.
286
‹¢ *+5
(
B. Only for 7• , when Qu&y precedes and Q2 follows:
Layer 1: Change of 7• to 3 (
y 39›*2«
by 8.3.17 n9n09Q•9Q^kuŠ
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q
7• -> 3 (
w 7•
pS (
Layer 2: Elision of 3 (
(
By 8.3.19 i9^5 2"$ÞŠ« covers the whole area to elide the 3 optionally.
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q
3(
7• ->i9^5
Op.
w 7•
pS (
(
8.3.22 ‹*i uËs"' «( gives *4e (compulsory) elision of 3 when ‹i ( follows.
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y (
/2 /s / ( (
Q
-> 3 ( i9^5
Op.7•i9^5
w 7•
pS (
6 0&1
E.g., o1u" Q•, Q—4y6 ƒu"S, 0&" 6 s.6
287
Qû"o to Layer 1: Change of 7• to ƒ

(
When 7• is after QT and followed by QT, ( 7• is replaced by ƒ, by 6.3.113 QT9 %9%‚ 6T"o‚ 6T1«.
(
When 7• is after QT and followed by ‹2, ( 7• is replaced by ƒ, by 6.3.114 ‹*2 S«.
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q ƒ ƒ
w -> 3 ( i9^5
Op.7•i9^5 7•
pS (
6 *+ by 6.1.87 w &5«.
With the preceding Q, 0& 6
(
When QT follows, y ^ *+ by 6.1.109 ‘™5 ô"t"o*T«
^ku–
^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q ,› ,
w -> 3 ( i9^5
Op.7•i9^5 7•
pS (
E.g., *2u9›‹', ( *2u9 u|5.
6
When %1Ý is not of 7• , none of these changes happen. E.g., ^4%*^, 6
^47*´5, 6
^4—y .
288
‹¢ *+5
C. For both % (and 7• , when % (follows:
Layer 1: elision of % (
By 8.3.14 %9 *%« ~ i9^5
^% QS ( ‹2 ( Ó% (
Q w/pS ( ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
% ( $/
^ku y /2 ( /s ( / (
Q
i9 % ((either % (or 7• )
w
^5
pS (
After the elision, if the preceding letter is Q&, ( it is elongated by 6.3.111 Oi9^1
6 %'T1, ‹%/ %Å5, 2fk %"—T1.
y 39›*2«. E.g., ^4"
n9n09Q•9Q^kuŠ
If the %1Ý is of 7• and preceded by Qu&y, changes for 7• take precedence. E.g., '49%‡5.
If the %1Ý is not of 7• , and preceded by Qu&y, i9^ of %1Ý by 8.3.14 %9 *% « takes place. E.g.,
6 %'T1.
^4"
( its ^ku y and Q2 as

The area not covered by any of the sūtras so far, which is pS as ( its
6 u, x*snyu*T.
^% remains as %.( E.g., x*s%`, x*s%"Æ1, 07%1
Putting all together:

^% QS ( ‹2 ( Ó% (
Q w/pS ( $/ ( ( » (/l ( T/‡

( Ó ( S/¶ ( ( ^/Ý
( (
^ku y /2 ( /s ( / (
Q ,› , (
op. op.
w -> 3 ( i9^5
Op.7•i9^5 2( s(
%(
i9 ≍ op. *u —y4/35 ≍
pS ( ^5
289
A. ô"t % ((of both 7• and %)(

1) ( % (+ Q2 (
pS +
6 u « x*snyu*T«
Ex. 07%1
1)-1 p, ƒ + % (+ % (
Ex. ‹%/ %Å5 « 2fk %"—T1 «
2) ( % (+ Ó% (
QS +
( : + Ó% (
QS + 8.3.15 Ó%u "439*uy —y4/35 « ~ ôŠ
2)-1 (
QS + ( Ó% (
+ 8.3.34 *u —y4/3Š 5 «
Ex. 4'Æ1 «
2)-2 ( ≍ + $6/^ 6
QS + 8.3.37 $6x95 ≍$ ≍^… S «
Ex. Ê≍Ó' «( ^4≍^
6 456 «
2)-3 ( 2+
QS + ( S/2
6 ( 6 š56 «
8.4.40 Æ95 š4"
Ex. %"'š «
2)-4 ( s+
QS + ( »6/s ( 8.4.41 U4
6 " U56 « ~ Æ95
Ex. %"'U/$T1 «
290
‹¢ *+5
B. ô"t (
becomes 7• by 8.2.66 6 7• :«
—s9
1) Q + 7• + Q
Q+ƒ+Q (
6.1.113 QT9 %9%‚ 6T"o‚ 6T1 « ~ ƒT Q*T
,› 6.1.87 w &5«,
6 6.1.109 ‘™5 ô"t"o*T« ~ ^ku5y
Ex. *2u9›Î5 «
2) Q + 7• + ‹2 (
Q + ƒ + ‹2 ( ( ƒT Q*T
6.1.114 ‹*2 S « ~ QT5 Q‚ 6T"T %95 (
, + ‹2 ( 6.1.87 w &5«
6
Ex. *2u9 u|5 «, '49%‡5 «
3) w + 7• + Q2 (
( Q2 (
w+3+ y 39›*2 « ~ %95
8.3.17 n9n09Q•9Q^kuŠ
3)-1 w+ Q2 ( 8.3.19 i9^5 2"$ÞŠ « ~ 3Š (*u$®5)
Ex. o1u" w0•*t «
3)-2 w+ ‹i ( 8.3.22 ‹*i uËs"' «( ~ i9^5 3Š (*u$®5)
6 0&1
Ex. 0&" 6 s6«
C. ô"t % ((%1Ý"t-QZ3"*4 such as ^4%

6 ,( #"T%,( and 5/1 and 6/1 of xot, such as $T%y6 )(
1) Q + % (+ Q
6
Ex. ^4%*^« (
(in contrast to *2u9›‹'«)
2) Q + % (+ p2 (
6
Ex. ^47*´5 « (in contrast to Q—4y6 ƒu"S«), ^4—y
6 « (in contrast to *2u9 u|5«)
3) Q + % (+ % (
Q+ %( 8.3.14 %9 *% « ~ i9^5
w+ %( y o/•Ì›&5«
6.3.111 Oi9^1 ^kuŠ
6 %'T1«(in contrast to '49%‡5«)
Ex. ^4"
291
D. ‘s5 and 5
1) ‘s5/ 5 + Q
‘s9›/ 9›
Ex. ‘s9›3'«,( 9›‹' «( (same as A. 1))
2) ‘s5/ 5 + w/pS (
Ex. ‘s ƒÎT1 « ƒÎT1 « (same as A. 3)-1)
3) ‘s5/ 5 + ‹i (
‘s/ + ‹i ( 6
6.1.132 ‘T§o95 i9^9›$9%4Ø '" 1 ‹*i «
Ex. ‘s *uŽ56 « *2u5«
292
‹¢ *+5
( -
[LSK] p*T 0 ^Í *+#$%&' 1/1
Thus ends the section of five *+.
Even though there are only three *+ sections in i•*6 }"t$…'o6 /, *+ is known to be
five in number in tradition. According to i•*6 }"t$…'o6 / *»å&/$"%, the five types of *+ are:
1. Qˆ *+5
2. ‹¢ *+5
3. *u 0y *+5
4. €"*o *+5
This is included in i•*6 }"t$…'o6 / under ‹¢ *+.
6
5. Q4€"% *+5
This is included in i•*6 }"t$…'o6 / under ‹¢ *+.
In * }"t$…'o6 /, the five *+s are told to be:

1. Qˆ *+5
2. Q*S +… #$| *Tn"u5
3. ‹¢ *+5
4. *u 0y *+5
5. €"*o *+5
Since #$| *Tn"u is about absence of *+, it is not counted as *+ by itself.
293
Index
Q 6
ƒô12›1 —44"* (
$ pT 1.3.2................................... 122
Q$5 u&Ë o/•y5 6.1.101....................................... 153 ƒ^ 0"y5 *¨3"3901 1.4.58 ..................................... 140
QSš 1.2.28 ......................................................... 53 ƒ^ 0"y”*T q"T… 6.1.91 ........................................ 142
QS9›•"*o *» 1.1.64 .......................................... 146 ƒ%& %•( ^%5 1.1.51 ...........................................50, 124
Q&*6 oT ( u&yŠ S"#e35 1.1.69 ............................. 85 ù
QT9 %9%‚ 6T"o‚ 6T1 6.1.113 ...................................... 267 ù$"i9› €o/•y‚ 6T5 1.2.27 ................................ 76
6
Q`"44"* y T 6 u" 8.3.2 .............................. 244
$5 ^kuŠ ‘
Qo2y4 i9^5 1.1.60 ............................................... 73 ‘™5 ô"t"o*T 6.1.109 ....................................... 155
6 1.1.2 .................................................. 116
Qo1™ ( 0&5 (
‘*™ ^%–^' 6.1.94 ............................................. 144
Q4*S S 8.4.47 ..................................................... 99 ‘S9›3u"3"u5 6.1.78 ........................................... 107
6
Q44"* (
$"T ^%9›4 6
€"%5 8.3.4............................... 245 6
‘T§o95 i9^9›$9%4È ( '" 1 ‹*i 6.1.132 ............ 280
6
Q4€"%Š 3*3 ^% u&y5 8.4.58............................. 224 ‘e1qek’ 6 6.1.89 .............................................. 133
Q4 1$"i ( *2T ( uyŠ 1.1.55.............................. 42, 160 ,
Qt"*ouí 6.1.85 ................................................ 151 ,'"™9š 6.1.95 ................................................ 149
Qi9›•Š 1.1.52 ........................................ 39, 103 $
Qu™( ê9»"34Š ................................................. 163 $"4"l1*¸T1 8.3.12 ................................................ 259
w $6x95 ≍$≍^… S 8.3.37 ...................................... 256
w*o%•14 ‹1T" 1.1.71 ......................................... 75 Ó
w*o*Èy»6¸u5 1.3.5 ................................................. 17 Ó%u "439*uy —y4/35 8.3.15 ................................ 246
wo15 ^%Š 1.1.54 ......................................... 41, 212 Ó*% S 8.4.55 ..................................................... 215
6 6.1.87 .................................................. 119
wo ( 0&5 ™
w±t… »*$T… 1.1.46 .................................... 48, 232 ™'9 •€"o*S ™'& (
6 *4e' (
8.3.32 ........................ 241
p *™í 1.1.53 ...................................................40, 161
6 }/ 1.1.3 .................................................. 51
p$9 0&u| &95 $6q6$ ( 2*% 8.3.28 ...................................... 233
p$9 3&*S 6.1.77 .................................................. 93 S
pŒ1 S ................................................................. 165 S»• 6 k• 1.3.7.............................................................. 19
ƒ ¶
ƒí •7o"§5 1.2.29 ................................................... 77 ¶1 S 6.1.73 ........................................................ 260
y 8.4.61 ................................. 210
ƒo5 v"Æf95 ^kuŠ ð
6
ƒô12›1 —44"* (
$ pT .............................................. 14 (
ð39 ‹9›ÕT%Š"' 8.4.62 .................................... 217
294
Index
ð%9 ð*% u&Ë 8.4.65............................................ 214 n

ði" —29›t1 8.2.39 ............................................ 203 nku"o39 q"Tu5 1.3.1 ............................................ 141
(
ði" —2 ð*2 8.4.53 ........................................... 101 y 39›*2 8.3.17........................... 270
n9n09Q•9Q^kuŠ
¸ '
¸5 * q»6 ( 8.3.29.................................................. 236 ( 1.1.47........................................ 49
*'oS9›•"T ^%5
ñ 'Ó 6
6 4"* $"uS49›44"* $5 1.1.8 ............................. 78
y o/•Ì›&5 6.3.111 ............................... 275
Oi9^1 ^kuŠ '9 %"*— '5 ò… 8.3.25 ........................................ 226
T 6
'9›4€"%5 8.3.23 ............................................... 221
T^%ÆB"iŠ 1.1.70 .......................................... 117 3
6
T¼"*oe§%Š 1.1.67 ............................. 34, 96, 211 (
3‡" ™(¤'4o16 25 '"4"' 1.3.10 ...................46, 109
y 1.1.66 ............................ 32, 95
T*¼*ü*T *4*oýU 1 ^kuŠ 6
3%9›44"* 6
$1 ›44"* $9 u" 8.4.45 .......................... 204
(
TŠ ^%'"l1*¸T' 8.1.2 ......................................... 258 314 *u*qÆotŠ 1.1.72........................................ 36
TŠ i9^5 1.3.9 ............................................... 21, 74 314 *u*qÆotŠ 1.1.72...................................... 111
TÞ (
6 "Š#3É u&y' 1.1.9 ....................................... 80 %
T95 *s 8.4.43....................................................... 202 %9 *% 8.3.14 ........................................................ 274
T9*iy 8.4.60 ........................................................ 208 6 8.2.69 .................................................... 273
%9› *^
o i
%"}kT 1 S.............................................................. 166 i2òT*}T1 1.3.8 .................................................. 20
4 i9^5 2"$ÞŠ 8.3.19 ........................................ 126
(
4 ô"t" 9%4"' 8.4.42 ......................................... 198 u
• 6 "5 1.3.4 ............................................. 16
4 *un´… T¼ u" ô"tŠ 8.4.59 .............................................. 225
4^%1 45 8.3.27 ..................................................... 231 u" 2*% 8.3.36 ..............................................248, 265
4š 8.3.30 .......................................................... 238 u"t9 *3 #e31 6.1.79 .......................................... 113
4š"ô"tŠ ð*i 8.3.24 ...................................... 222 (
*u#*Ts1q 1 ^% $"3y' 1.4.2 ...................................... 277
(
4Ÿ¶Z#2"4 8.3.7 ............................................... 253 *u —y4/3Š 5 8.3.34 .................................247, 264
6
4/S •%4o"§5 1.2.30 ................................................. 77 (
u|*}%"o•S 1.1.1 .................................................... 130
( 8.3.10 ........................................................ 255
4Q4 ^1 u|*}%1*S 6.1.88 .................................................... 131
^ 2
ô"t"Ò" 6.1.76 ................................................... 261 2Ÿ¶9›*» 8.4.63 ................................................. 219
^%5 *ü$s y5 *‹T" 1.4.109 .................................... 88 6 ( 8.3.31 ..................................................... 239
*2 T$
6 Ó[†%1 8.3.6 ............................................... 251
^'5 s
(
^ku`y "* }' 8.2.1 .................................................. 128 s5 #e3Š 1.3.6 .................................................. 18
295
sé/ v"4 1390" 1.1.49........................................ 31, 94 (u"*Ty$')( 3ui^%1 3ui" u" ................................... 229
(u"*Ty$')( 3*¼*µ*qÆo"o"u¢œ‹&1...................37, 112
390"tŠ i9^5 8.2.23 ....................................... 102 (u"*Ty$')( 2$˜"*os 6 ^%–^ u"Î' .......................
( 147
6 8.3.5 .................................................... 243
'5 *» (u"*Ty$')( ^$"4"
6 9 u´Z5 ................................ 249
'"‹"%5 €*%T5 1.2.31........................................... 78
uy` *un"s" 095 6.1.122 ...................................... 157 `k ¨'5
6 7• : 8.2.66 ................................................ 266
—s9 (
1.1.1 u|*}%"o•S .................................................... 130
6
*â™t (
ô' 1.4.14 .............................................. 90 6 ................................................. 116
1.1.2 Qo1™ ( 0&5
(
9›*S i9^1 S1°"o^k%&' 6.1.134 ........................... 283 6 }/.................................................. 51
1.1.3 p$9 0&u|
v"4 1›t%T'5 1.1.50 ....................................... 44, 97 1.1.46 w±t… »*$T… ....................................48, 232
‹ ( ........................................ 49
1.1.47 *'oS9›•"T ^%5
(
‹i•' 1.3.3 ................................................. 15, 71 1.1.49 sé/ v"4 1390" ........................................31, 94
(
‹*i uËs"' 8.3.22 ............................................... 272 1.1.50 v"4 1›t%T'5 .......................................44, 97
‹i9›4t%"5 3905 1.1.7 ....................................... 89 1.1.51 ƒ%& %•( ^%5 ...........................................50, 124
‹*2 S 6.1.114 .................................................... 269 1.1.52 Qi9›•Š ........................................39, 103
‹1 '^%1 u" 8.3.26 .................................................. 228 1.1.53 *™í ...................................................40, 161
1.1.54 wo15 ^%Š..........................................41, 212
^*%n"s"
1.1.55 Q4 1$"i ( *2T ( uyŠ ..............................42, 160
6 %9§% Ri/35
(^*%n"s") ^ku^y %-*4e-Qt%¹-Qû"o"4"'§
1.1.60 Qo2y4 i9^5............................................... 73
...................................................................... 279
1.1.64 QS9›•"*o *» ......................................... 146
u"*Ty$"*4
y ............................32, 95
1.1.66 T*¼*ü*T *4*oýU 1 ^kuŠ
(u"*Ty$')( Q“" *‹Õ"'^6 û"4' ............................
( 136
6
1.1.67 T¼"*oe§%Š ..............................34, 96, 211
(u"*Ty$')( Qè^*%'"&1 S ....................................... 115
1.1.69 Q&*6 oT ( u&yŠ S"#e35 ............................ 85
(u"*Ty$')( Q4"Äu*T40%/&"*'*T u"Î' ..................
( 200
1.1.7 ‹i9›4t%"5 3905....................................... 89
(u"*Ty$')( x{u&y39*'y‡5 "uÁ3Ñ u"Î' ....................
( 81
1.1.70 T^%ÆB"iŠ ......................................... 117
(u"*Ty$')( xT1 S T|T/3" '" .................................
1 138
1.1.71 w*o%•14 ‹1T" ........................................ 75
(u"*Ty$')( 093kTy … ¶¡Š^6 û"4' ...........................
( 115
1.1.72 314 *u*qÆotŠ.................................36, 111
(u"*Ty$')( S39 *ÒT/3"5 2*% ^…ï% "o1*%*T u"Î' .....
( 234
1.1.8 'Ó 6
6 4"* $"uS49›44"* $5 ............................. 78
(u"*Ty$')( ¶•' Q'/*T
( (
u"Î' ............................... 220 (
6 "Š#3É u&y' ......................................
1.1.9 TÞ 80
(u"*Ty$')( #e31 n"s"3" *4e' ...............................
( 206
1.2.27 ù$"i9› €o/•y‚ 6T5 ................................ 76
(u"*Ty$')( #u„T%$¬iu 4"&yo2"4"'|& 1 ............... 139
1.2.28 QSš ........................................................ 53
(u"*Ty$')( #" ‹9ñ9N1s •³1s 6 ..................................... 136
1.2.29 ƒí •7o"§5 .................................................. 77
(u"*Ty$')( 3&5 #*Ts1q9 u"Î5 ................................ 104
6
1.2.30 4/S •%4o"§5 ................................................ 77
296
Index
1.2.31 '"‹"%5 €*%T5........................................... 78 6.1.89 ‘e1qek’ 6 .............................................. 133

1.3.1 nku"o39 q"Tu5 ............................................. 141 6.1.91 ƒ^ 0"y”*T q"T… ........................................ 142
(
1.3.10 3‡" ™(¤'4o16 25 '"4"' ................... 46, 109 (
6.1.94 ‘*™ ^%–^' ............................................. 144
6
1.3.2 ƒô12›1 —44"* (
$ pT ............................. 14, 122 6.1.95 ,'"™9š ................................................ 149
(
1.3.3 ‹i•' ................................................. 15, 71 y o/•Ì›&5 ............................... 275
6.3.111 Oi9^1 ^kuŠ
• 6 "5 ............................................. 16
1.3.4 4 *un´… T¼ (
8.1.2 TŠ ^%'"l1*¸T' ........................................ 258
1.3.5 w*o*Èy»6¸u5 ................................................. 17 (
8.2.1 ^ku`y "* }' ................................................. 128
1.3.6 s5 #e3Š ................................................... 18 8.2.23 390"tŠ i9^5 ...................................... 102
1.3.7 S»• 6 k• .............................................................. 19 8.2.39 ði" —29›t1 ........................................... 203
1.3.8 i2òT*}T1 ................................................... 20 8.2.66 6 7• : ............................................... 266
—s9
1.3.9 TŠ i9^5 ............................................... 21, 74 6 .................................................... 273
8.2.69 %9› *^
1.4.109 ^%5 *ü$s y5 *‹T" .................................... 88 8.2.84 %"}kT 1 S .................................................. 166
6
1.4.14 *â™t (
ô' .............................................. 90 ( ....................................................... 255
8.3.10 4Q4 ^1
(
1.4.2 *u#*Ts1q 1 ^% $"3y' ....................................... 277 8.3.12 $"4"l1*¸T1................................................ 259
1.4.58 ƒ^ 0"y5 *¨3"3901 ...................................... 140 8.3.14 %9 *% ........................................................ 274
6.1.101 Q$5 u&Ë o/•y5....................................... 153 8.3.15 Ó%u "439*uy —y4/35 ................................ 246
6.1.109 ‘™5 ô"t"o*T ....................................... 155 y 39›*2........................... 270
8.3.17 n9n09Q•9Q^kuŠ
6.1.113 QT9 %9%‚ 6T"o‚ 6T1 ...................................... 267 8.3.19 i9^5 2"$ÞŠ ........................................ 126
6.1.114 ‹*2 S .................................................... 269 6
8.3.2 Q`"44"* y T 6 u".............................. 244
$5 ^kuŠ
6.1.122 uy` *un"s" 095 ...................................... 157 (
8.3.22 ‹*i uËs"' .............................................. 272
6.1.123 Qu™( ê9»"34Š .................................... 163 6
8.3.23 '9›4€"%5 ............................................... 221
6.1.124 pŒ1 S .................................................... 165 8.3.24 4š"ô"tŠ ð*i ..................................... 222
6
6.1.132 ‘T§o95 i9^9›$9%4È ( '" 1 ‹*i ............. 280 8.3.25 '9 %"*— '5 ò… ........................................ 226
(
6.1.134 9›*S i9^1 S1°"o^k%&' ........................... 283 8.3.26 ‹1 '^%1 u" ................................................. 228
6.1.73 ¶1 S ......................................................... 260 8.3.27 4^%1 45 .................................................... 231
6.1.76 ô"t"Ò" ................................................... 261 8.3.28 &95 $6q6$ ( 2*% ...................................... 233
6.1.77 p$9 3&*S .................................................. 93 8.3.29 ¸5 * q»6 .................................................
( 236
6.1.78 ‘S9›3u"3"u5 ............................................ 107 8.3.30 4š ......................................................... 238
6.1.79 u"t9 *3 #e31........................................... 113 8.3.31 *2 T$ (
6 ..................................................... 239
6.1.85 Qt"*ouí ................................................ 151 8.3.32 ™'9 •€"o*S ™'& (
6 *4e' (
........................ 241
6 .................................................. 119
6.1.87 wo ( 0&5 8.3.34 *u —y4/3Š 5 .................................247, 264
6.1.88 u|*}%1*S ..................................................... 131 8.3.36 u" 2*% ..............................................248, 265
297
8.3.37 $6x95 ≍$≍^… S ....................................... 256 8.4.55 Ó*% S ..................................................... 215

6
8.3.4 Q44"* (
$"T ^%9›4 6
€"%5 ............................... 245 6
8.4.58 Q4€"%Š 3*3 ^% u&y5 ............................ 224
6 .................................................... 243
8.3.5 '5 *» 8.4.59 u" ô"tŠ .............................................. 225
6 Ó[†%1 ............................................... 251
8.3.6 ^'5 8.4.60 T9*iy ....................................................... 208
(
8.3.7 4Ÿ¶Z#2"4 ............................................... 253 y ................................ 210
8.4.61 ƒo5 v"Æf95 ^kuŠ
(
8.4.42 4 ô"t" 9%4"' ......................................... 198 (
8.4.62 ð39 ‹9›ÕT%Š"' .................................... 217
8.4.43 T95 *s ....................................................... 202 8.4.63 2Ÿ¶9›*» ................................................. 219
6
8.4.45 3%9›44"* 6
$1 ›44"* $9 u" ........................... 204 8.4.65 ð%9 ð*% u&Ë........................................... 214
8.4.47 Q4*S S ..................................................... 99
(
8.4.53 ði" —2 ð*2 ........................................... 101
298
www.arshaavinash.in
WEBSITE FOR FREE E-BOOKS ON VEDANTA & SANSKRIT
Pujya Swami Dayananda Saraswati launched Arsha Avinash Foundation’s

website www.arshaavinash.in on Dec 31, 2014.
All the E-books available on the website can be downloaded FREE!
PUJYA SWAMI DAYANANDA SARASWATI- A BRIEF BIOGRAPHY BY N.

AVINASHILINGAM. It is available in English, Tamil, Hindi and Portuguese.
SWAMI PARAMARTHANANDA’S TRANSCRIBED CLASS NOTES: Available class

notes are Gita (3329 pages), Isavasya Upanisad, Kenopanisad, Kathopanisad, Prasna
Upanisad, Mundaka Upanisad, Mandukya Upanisad with karika, Aitareya Upanisad,
Chandogya Upanisad, Brihadarnyaka Upanisad (1190 pages), Brahma Sutra (1486
pages), Tattva Bodha, Atma Bodha, Vivekachudamani (2038 pages), Panchadasi,
Manisha Panchakam, Upadesha Saara, Saddarsanam, Jayanteya Gita, Jiva Yatra,
Dhanyastakam, Advaita Makaranda, Dakshinamurthy Stotram, Drg Drsya Viveka and
Naishkarmya Siddhi.
BRNI MEDHA MICHIKA’S BOOKS ON SANSKRIT GRAMMAR: Enjoyable

Sanskrit Grammar Books- Basic Structure of Language, Phonetics & Sandhi,
Derivatives (Pancavrttayah), Dhatukosah, Astadhyayi, Study Guide to Panini Sutras
through Lagu Siddhanta Kaumudi, Sanskrit Alphabet Study Books- Single Letters,
Conjunct Consonants.
There are many more books and articles on Indian culture and Spirituality, Chanting,
Yoga and Meditation. There are also books in Tamil on Vedanta.

104 Third Street, Tatabad, Coimbatore 641012, India
Phone: +91 9487373635
E mail: [email protected]
www.arshaavinash.in

Laghu 1

Uploaded by

Copyright:

Available Formats

Laghu 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Laghu 1

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What are some of the other books available on the website?

What are some of the other books available on the website?

The Study Guide to

Laghu Siddhanta Kaumudi

Part 1 – Sajna Prakaranam &

E Book published by:

Arsha Avinash Foundation

Electronic version of this book is available at:

Printed version of this book is available at:

This book is the first volume in a series titled

Purpose of this book

This book is a study guide to Pā ini-Sūtra through i•*6 }"t$…'o6 /.

1. Q*q$"%/ – Who should read this book:

2. *us35 – Subject matter of this book:

3. #39—4' –( Usefulness of studying this book:

The connections are between:

The Sanskrit word for Pā inian grammar is ^"*&4/3Z"$%&'. (

^Th*i5 (150 BCE)

Pā ini’s QU"j"3/ `k ^"l5

`k #n1o"5 (Types of Sūtras)

Sūtras in QU"j"3/ can be categorized into six types.

2) !/ (named, meaning of the name)

3) !"-#o125 (usage of !")

Following Pujya Swamiji’s guidance, we shall commence our learning of Pā ini

1.3.6 s5 #e3Š « #e3Š 8

This is the first sūtra of pT-( !" section.

[ !" `k ']( 1.3.2 6 $ pT «(

[LSK] ƒ^o12 1 7/1 Q44"*

The following śloka tells what are considered to be ƒ^o1252

[ !" `k ']( 1.3.3 ‹i•' «( ~ ƒ^o12 1 pT (

‹i 1/1 ( « ~ ƒ^o12 1 7/1 pT 1/1

2 words in the `k ; 2 words as Q4u|6 *§

ƒo"‹%&"*4 (examples) are:

[*4s1q `k ']( 1.3.4 4 *un´… T¼ ( ( Q•' (

3 words in the `k ; 4 words as Q4u|6 *§

Tu0y (T 6 S ( ' S). Total seven letters.

[LSK] *un*´v"5 1/3 Tu0y- -'"5 1/3 4 0 pT5 1/3 «

[ !" `k ']( 1.3.5 (

ƒo"‹%&"*4 (examples) are found only in q"Ts6 :

[ !" `k ']( 1.3.6 (

ƒo"‹%&"*4 (examples) are found only in #e3s:

[ !" `k ']( 1.3.7 (

• #e3Š 6/1 – Suffix; in ¬+sé/ to w*o5.

[LSK] #e3"o… 7/1 S»6 k 1/2 pT… 1/2 Æ5 III/2«

ƒo"‹%&"*4 (examples) are found only in #e3s:

[ !" `k ']( 1.3.8 (

[LSK] T*}T-u—y-#e3"±"5 1/3 i2$u0"y5 1/3 pT5 1/3 Š56 III/3«

ƒo"‹%&"*4 (examples) are found only in non-T*}T #e3s:

After definitions of pT, ( this *u*q `k teaches that pT should

[*u*q `k ']( 1.3.9 TŠ i9^5 «

TŠ 6/1 i9^5 1/1 ।

2. Q*q$"% `k ' ( (Sūtra which tells the topic of following sūtras)

• 3.1.1 #e35 1/1 « up to 5.4.160

• 3.1.2 ^%š 1/1, 0 « up to 5.4.160

• 4.1.76 T*}T"5 1/3 « up to 5.4.160

• 6.4.1 Q¹Š 6/1 « up to 7.4.97

• 6.4 129 nŠ 6/1 « up to 6.4.175

[4s1q `k ']( 1.3.4 4 un´… T¼ ( ( Q•' (

[LSK] un´v"5 1/3 Tu0y- -'"5 1/3 4 0 pT5 1/3 «

ƒo"‹%&"4 (examples) are found only in non-T}T #e3s:

After definitions of pT, ( this uq `k teaches that pT should

[uq `k ']( 1.3.9 TŠ i9^5 «

1st 2.1.1 '" 5 3.1.1 #e35 4.1.1 #"T^o$"T (

3.1.91 q"T95 4.1.76 T}T"5 6.1.72 ‹T"3"' (

3rd 1.3.2 pT ( 2.3.1 un´5

What uq `k s enjoin are:

B. Interpretations of words in '"4"q$%& (Totuq and To"ouq)

1.1.72 314 uqÆotŠ « Supply “Qt” to *u21s&.

C. Deciding v"44, ( when there are many v"44s(

( QS 7/1 « ~ ‹T"3"' 7/1

( pT 0 ƒ§%Š 6/1 « ~ 4*oýU"T 5/1

B. Interpretations of words in '"4"q$%& (Totuq and To"ouq)

• 1.1.72 314 uqÆotŠ «

[^%n"s" `k ']( 1.1.72 314 u*qÆotŠ « ~ €Š –^Š

314 3/1 uq5 1/1 TotŠ 6/1 « ~ €Š 6/1 –^Š 6/1

( uq5 1/1 To"o… 7/1 Qi-œ‹&1

C. Deciding v"44, ( when there are many v"44s(

[^%n"s" `k ']( 1.1.53 ™í « ~ sé/ Qi5 Q•Š

[^%n"s" `k ']( 1.1.55 Q41$"i ( 2T ( uyŠ «