Sorry, your browser does not support JavaScript!
Sanskrit Grammar
Beta version only through Ch. 4 Pronouns, Cardinals & Ordinals -- as of Feb 11, 2018
Invocation Acknowledgements Abbreviations Contents Introduction
Ch. 1 The Alphabet Ch. 2 Rules of Sandhi Ch. 3 Declension Ch 4 Pronouns and Numerals Ch. 5 Conjugation of Verbs Ch. 6 Formation of Words Ch. 7 Syntax
Glossary Bibliography Index

Format by A.K. Aruna, 2018 ver.1.0: UpasanaYoga. The contents here is taken from the author's book The Aruna Sanskrit Grammar Reference the first of a five book series available online in pdf format. If downloaded, requires installed Devanāgarī Siddhanta1.ttf font, downloadable from UpasanaYoga. If run from UpasanaYoga website, it alternatively can use online Web Font.
The behavior of actions are device, browser, and browser-settings dependent. As of now, Google's Chrome browser works fast and fully. Future versions of HTML language or JQuery libraries also may stop supporting some of the features employed here.

by A.K. Aruna
First Print Mar 2011 (ISBN 978-0-9818640-0-6)
First Update to HTML Mar 2018 with Creative Commons International License:
This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. To view a copy of this license, visit, or click the following logo:
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License

नारायणं नमस्कृत्य नरं चैव नरोत्तमम्।
देवीं सरस्वतीं चैव ततो जयमुदीरयेत्॥

Nārāyaṇaṃ namaskṛtya naram caiva narottamam.
Devīṃ sarasvatīṃ caiva tato jayam udīrayet.

Bowing to Lord Nārāyaṇa (Kṛṣṇa), and to Nara, the best of men (Arjuna), and to the Goddess (of knowledge) Sarasvatī, may one now commence the (epic) Jaya (‘Victory’).

Go to Top ('Home' key on many Windows browsers)

Abbreviation Meanings
a. or adj.adjective(s)
aor.aorist (tense)
॰-as first member
-॰as second member
आ॰ or Ā.Ātmanaipada
changes to
conj. cons.conjunct consonant(s)
f. or fem.feminine(s) (vowel or consonant)
fut. pt.future participle (active)
in. pt.indeclinable participle
init.initial (letter)
m. or masc.masculine(s)
med.medial (vowel)
mid.middle (Ātmanaipada)
n. or neut.neuter(s)
पर॰, Par. or P.Parasmaipada
past act. pt.past active participle
pp.past participle (passive)
perf.perfect (tense)
poss. prn.possessive pronoun
pot.potential (mood)
pot. ps. pt.potential passive participle
pr. pt.present participle (active)
pr. mid. pt.present middle participle
prn. a.pronominal adjective
redup. perf. pt.reduplicated perfect participle (active)
s.f.stem final
w/wowith or without

Go to Top ('Home' key on many Windows browsers)




Using this book

Introduction to Sanskrit

The origins of Sanskrit

Vedic Sanskrit versus Classical Sanskrit

If Panini wrote the definitive grammar book, then why are we offering this work?

Why is this series better for learning Sanskrit? And to whom is this work intended?

Why is this series better for learning Sanskrit? And to whom is this work intended?

The Aruna Sanskrit Language Series

Ch. 1 The Alphabet

The Sounds in the Sanskrit Language

Devanāgarī alphabet

Conjunct consonants


Ch. 2 Rules of Coalescence - सन्धि

Vowel strengthening

General सन्धि

Vowel सन्धि

Permitted final consonants

विसर्ग (:) सन्धि

Consonant सन्धि

Internal सन्धि

Ch. 3 Declension

Normal case terminations

Accent in declension

Consonant-ending unchangeable stems

Consonant-ending changeable stems

Two-form stems

Stems in अत्, मत्, वत्

Stems in (ई)यस्

Stems in इन्, मिन्, विन्

Three-form stems

Stems in वस्

Stems in अन्

Stems in मन्, वन्

Stems in अच्

Vowel-ending noun stems

Stems in अ, आ

Stems in इ, उ

Stems in ई, ऊ

Stems in ऋ

Stems in ऐ, ओ, औ

Ch 4 Pronouns and Numerals


Personal अहम्, त्वम्

Common Pronominal terminations

Relative proximity of pronouns

Demonstrative (ए)तद्, इदम्, अदस्, एनद्

Relative यद्

Interrogative किम्



Pronominal adjectives

Compound pronouns



Numeral adverbs and derivatives

Ch. 5 Conjugation of Verbs

Overview of conjugation

Ten classes of roots

The four verbal base tenses and moods

Reduplication rules

Irregular verbal bases and forms

Perfect tenses

Aorist tense

Benedictive mood

Future tenses

Conditional mood




Intensive or frequentives


Ch. 6 Formation of Words

Parts of speech


Verbal compounds

Independent prepositions

Prepositional adverbs, participles, nouns

Adverbs formed with suffixes

Conjunctive and adverbial particles

Nominal stem formation

Bare roots

Comparative & superlative suffixes

Primary suffixes

Secondary suffixes


Nominal Compounds

Compound types and accent

Oblique case तत्पुरुष

Negative नञ्-तत्पुरुष

Same case कर्मधारय

Numerical द्विगु

Prepositional प्रादि

Indeclinable गति

Non-independent उपपद

Adjectival बहुव्रीहि

List द्वन्द्व

Indeclinable-like अव्ययीभाव

Ch. 7 Syntax

Prose sentence order


Definite and indefinite articles




Nominative case

Accusative case

Instrumental case

Dative case

Ablative case

Genitive case

Locative case

Locative and genitive absolute


Tenses and Moods

Present tense

Past tense

Future tense

Imperative mood

Potential mood

Benedictive mood

Conditional mood

Passive and impersonal construction



Intensive or frequentives





Go to Top ('Home' key on many Windows browsers)

Using this book

This book provides the succinct grammar rules and structures for the Sanskrit language. If you are unfamiliar with Sanskrit grammar and wish to learn how to understand and apply these rules and structures for reading Sanskrit, please get the companion book, The Aruna Sanskrit Grammar Coursebook: 64 Lessons Based on the Bhagavad Gita Chapter Two by this author. The Aruna Coursebook and the Grammar Reference are designed to work together and are fully cross-referenced. To start this course of learning to read Sanskrit, begin with the Aruna Coursebook, which will assign specific sections of rules from this grammar book. In following all the lessons in the Aruna Coursebook, you will cover the entire grammar in this book and be able to fluently read with understanding the second chapter of the Bhagavad Gita in its original Sanskrit.

To go even further into the study of Sanskrit so that you can completely read and understand the entire Bhagavad Gita, get the full set of The Aruna Sanskrit Language Series, of which this Grammar Reference and the Aruna Coursebook are the initial texts. The design and benefit of this series will be fully explained below within the context of Sanskrit and the methods of learning this language using The Aruna Sanskrit Language Series. This series is discribed below.

Introduction to Sanskrit

Sanskrit is a living language with ancient Vedic roots. Here I am using the term “living language” in its obvious sense – that it is still existing, still spoken, still studied, still communicative (and not just descriptive of what once was), and at least in one most highly desirable and non-academic area of study, namely, the Vedanta studied by people who newly come to its study to this day in the hundreds or thousands every year, is an irreplaceable medium for its full appreciation.

Sanskrit is a living language with ancient Vedic roots. Here I am using the term “living language” in its obvious sense – that it is still existing, still spoken, still studied, still communicative (and not just descriptive of what once was), and at least in one most highly desirable and non-academic area of study, namely, the Vedanta studied by people who newly come to its study to this day in the hundreds or thousands every year, is an irreplaceable medium for its full appreciation.

Recent archeological and related studies, just now scratching the surface, are seeing some indications of what could be described as the Vedic culture in the Indus-Saraswati civilization at its first urban development peak between 2600 to 1900 B.C. and its antecedent development in the same area starting at least eight thousand years ago including the large settlement of Mehrgarh (168 acres, one-quarter sq. mi., in 5000 B.C. – by many times the largest Neolithic settlement of its time in the world). The sites linked to this civilization over time are spread over one million square kilometers (the largest of any ancient civilization in the East, West or New World) – from eastern Afghanistan to the Ganges River and from the Himalaya mountains to the Godavari River in West-central India (an area nearly as large as the states of California and Texas together), with over 2,600 sites and more being found on a regular basis.

Unlike other cultures and civilizations that have come and gone within history, this civilization has been amazingly stable for an enormous period of time and has developed a very rich and deep tradition. There naturally occurred population shifts over this vast time period. Its first great cities, such as Mahenjo-Daro on the Indus River, Harappa on the Ravi River, and others were abandoned by 1900 B.C. Because the Indian sub-continent, due to continental drift, was slowly colliding into the Asian continent, there were massive land displacements and several of the rivers frequently shifted course, leaving barren once fertile banks. The greatest problem, however, was the complete drying up of the major Saraswati River along whose banks the largest bulk of the settlements have been found – likely due to the redirection of its feeder rivers as a result of the reshaping of the Himalayan headwater landscape. In addition there was a desertification of this vast area over time with prolonged drought between 2200-1900 B.C. throughout the whole of West and South Asia.

As a result, the portion of this urban-based civilization within the affected area over time, family by family, picked up its roots and migrated mostly into the fertile Ganges River plains in village-size settlements, bringing their cultural riches with them. These people, along with the portions of the civilization outside of this affected area, eventually emerged transparently into the India of today. With other ancient cultures, displacements of this magnitude would have destroyed the central culture, but, with its traditions maintained by strong, individual families dedicated to preserving its cultural treasures, their core literature has survived to this day, and is many times more extensive than that of any other ancient civilization.

The traditions of this civilization has been and still are being passed on to succeeding generations through oral transmission, and recently now augmented with printed text. The earliest form of this tradition that has survived is its collection of sacred hymns, which originally existed as separate hymns scattered among many families who had preserved them. These hymns were subsequently compiled and edited into the four books of the Vedas. That compilation is traditionally held to be done by a group of scholars headed by a man called Vyasa. Vyasa is also traditionally ascribed to be the author (maybe the compiler, editor and contributor of the various stories) of the great epic called the Mahabharata (or at least the initial stage of the epic – likely called Jaya meaning “Victory”). Within this epic is the jewel called the Bhagavad Gita, a work of 700 verses consisting of a dialogue between the Lord-incarnate Krishna and the warrior prince Arjuna at the start of a great war. Vyasa is also a participant in the Mahabharata story, and his name is mentioned twice in the Bhagavad Gita itself.

The Mahabharata centers on a great war said to have been fought thirty-five years before the beginning of the Kaliyuga, the last and declining age of the four cyclical ages. The Kaliyuga commenced by one calendar in 3102 B.C. Elsewhere, with support in the Puranas, the war was said to have been fought in 2449 B.C. A third view expressed in the Puranas “assumes that a total of 1,050 years elapsed between the time of Parikshit (Arjuna’s grandson) and Mahapadma Nanda’s rule,” which would place the war around 1500 B.C. This later date coincides with the recent datings of the underwater ruins of the ancient port city of Dwaraka (1700-1600 B.C.) from where Krishna came; somewhat coincides with B. B. Lal’s archeological dating of the apparent oldest layer of Hastinapura, a central city of the Mahabharata, to the Painted Grey Ware phase between 1100 and 800 B.C.; and also generally fits with the post-resettlement period of the Vedic people moving from the Saraswati River, which had dried up by 1900 B.C., eastward to the fertile Ganges River basin, the geographical site of the Mahabharata story. The Mahabharata, like the Puranas, is a popularization of the teachings of the Vedas and the Vedic traditions, and the Bhagavad Gita best summarizes that vision of the core teaching.

This set of books, called The Aruna Sanskrit Language Series, revolves around the Bhagavad Gita to teach the Sanskrit language and to expose the student to the core tradition of the culture in which this language flourished. The Bhagavad Gita is the most accessible prime example of the depth of cultural heritage that has allowed Indian culture to survive many times longer than any other culture of the past or present.

The origins of Sanskrit

Many scholars, even till today, are slow in accepting the Indus-Saraswati civilization, also called the Harappian or the Indus Valley civilization, as the oldest known appearance of Vedic culture or Vedic literature and language. These things take time to be assimilated and accepted. Many of these scholars are beholding to a great deal of investment by themselves, their teachers, the institutions they work under, their funding sources, their nation’s empire builders, and even their religion in the so-called Aryan invasion theory and its varients, which claim the soon-to-be authors of the Vedas invaded, or at least migrated, into the Northwest Indian sub-continent around 1500 B.C. from parts unknown, but widely speculated from anywhere except the Indian sub-continent itself.

Now indeed there is, so far, little direct physical evidence that the Indus-Saraswati civilization was in fact Vedic; however, there is no direct evidence that it was not Vedic. The reason it appears to be Vedic is that there is at least some exemplary evidence for it, it can as well alternatively explain away the minimal and now largely discredited evidentiary claims of the invasion theorists, and it better fits what the Veda and supporting Purana and Epic literature itself tell us.

Who better to turn to than the authors of the main evidence in question? The many authors within the Veda, with little dispute by scholars, describe the homeland of their ancestors as none other than the Indus-Saraswati area, and they make no claims of them or their ancestors having invaded from or lived elsewhere. Plus the Purana and Epic literature confirm and in fact voluminously expound on this one homeland theme, quite out of tune with the supposed nomadic Aryan invaders, who were the imagined authors of the Vedas by the theorists. Some may claim that the Purana and Epic literature is myth, but they are myth intermeshed with historical fact. For example, Dwaraka, the supposed legendary home of our story’s hero Krishna said to have been submerged shortly after Krishna’s death in the Mahabharata, was recently found underwater extending out 700 meters off shore of present day Dwaraka. This ancient city was a thriving sea-port on the broad delta near where the Saraswati River would have emptied into the Arabian Sea. Its underwater ruins and relics date to 1700-1600 B.C. This date has to be reconciled with the fact that the archaeological evidence shows that Dwaraka, the home of the eminent “Arya” Krishna, was a late development of the Indus-Saraswati civilization, that the city is located over 600 km south from the center of the civilization around the middle of the Saraswati river, that the Sanskrit spoken by Krishna in the Mahabharata is a much later development from the Sanskrit of the Vedas, and also that Dwaraka is a “recent” city first mentioned only in the Mahabharata, which postdates the Vedas, the Ramayana epic and many of the Puranas.

Now these are people natively gifted with tremendous memories. That is why we still have a vast amount of their literature today after many thousands of years, and better preserved than any comparable literature from anywhere else in the world that is more than even a couple hundred years old. Moreover, the vast bulk of their literature is spent in tracing their origin back to its divine source, so they were far from indifferent to recounting their past. Greek ambassadors to India told the Greek historians Pliny and Arrian of a list of 154 Indian kings reported to go back to 6676 B.C. Moreover, the archeologically supported population shift from the Saraswati River area to the Ganges River area is also clearly reflected in the differences of the geographic descriptions present in the Veda from those in the later Brahmana and Purana literature. This shows that their literature did indeed reflect their history. That these people could have collectively forgotten their roots, forgotten from wherever else they supposedly came from, within, by most scholars reckoning, a matter of a few hundred years from the supposed invasion (1500 B.C.) and the supposed time (1200 B.C.) of the initial compositions of the Veda literature, is itself a self-contradicting, if not preposterous, supposition that the invasionist scholars must swallow.

Admitting that the old archeological studies, or rather their interpretations, used to prop up the invasion theory are shaky, some current scholars now rely on the matching old linguistic studies to keep the origin of the Vedic culture, or at least the source of the Vedic language, out of the geographical area of the Vedas. Linguistics shows how languages may be interlinked in their developments or may be broadly categorized as belonging to agrarian or pastoral cultures, but the science of linguistics has no tools to address the problem of locating any language in a geographical area at a point in time – only archeology can do this. Linguistics at best can be used to rationalize an already assumed geographical dispersion of a group, or family, of languages.

Now, if human beings never move from their homeland and if language developed in only one place on earth and spread neighbor to neighbor from there, then linguistics alone, with knowledge of the geographical constraints within the area in question, could offer theories to where and how the language might have dispersed over time – but such is not the case. Human history is far too complicated for linguistic theory to claim the equal of a single concrete archeological fact. It should not be surprising then that by geographically locating the Vedic language and culture in the Indus-Saraswati area a couple, or several, thousand years before 1500 B.C. and given the convincing facts that there were extensive contacts by sea and over land between these people and the rest of the old world, that this gives more than enough time for linguistic theory to equally, or better, explain how that civilization’s language influenced and was influenced by other languages scattered throughout Central Asia, Middle East and Europe – especially since the archeological evidence of these other languages in other areas date so far at the earliest to around 3000 B.C., allowing more than enough time for this interaction to have taken effect by then.

The most likely scenario, which will forever remain unproven because systematic writing appears to be a recent invention and an ephemeral relic, is that groups of humans have separately initiated oral languages, some more complex than other, for tens, if not hundreds, of thousands of years, and have been exchanging these languages, in part and in total, through inter-marriage, migration, conquest, travel, trade and other cultural contacts from pre-historic times. That there is a similarity between languages within a geographically linked area as Euro-Asia should surprise no one, nor should it prove that there then had to have been some one mysterious nomadic tribe sprinkling its sole invention to the illiterate masses. Such an adamant fancy for the belief in a single, literate nomadic tribe befits only one who has an ulterior motive for its promulgation, as has been apply evidenced and admitted in the case of India by the European colonizers of India, the original inventors of the theory.

The current reluctance of some Western and Indian scholars of today to impartially re-appraise the frailty and self-promoting circularity of the theory may rest not only on the investment reasons mentioned above, but even on their vehemence, for whatever reason, against what many call “Hindutva,” the current revival of the core religion of India – viewing the loss of this theory as some kind of victory for the so-called “Hindutva crowd.”

Such is no excuse for not coming to grips with the facts as they present themselves. The fact is, the Vedas, the Puranas and the Epics, which form the basis of what they called Sanatana Dharma and is now called Hinduism, uniformly present their culture and language as beginningless. That the core of this tradition, the Vedas, are maintained unmanifest through the mechanism of karma and later reintroduced to mankind (once they have the language to receive it) in succeeding creation cycles each lasting billions of years, again and again – in current terminology, big-bang after big-bang. Those who appreciate and own up to this heritage would see no “victory” in pushing back this cultural heritage’s so-called origins just a few thousand years within this present cycle. Religious and anti-religious sentiments, as well as politics, indeed play a role in directing or misdirecting the search for historical truth, but not in establishing the truth.

A more fruitful and less speculative and contentious approach in this area may be to look at the movement of peoples within this region in the distant past, since migration and inter-marriage would have played at least some concrete role in the movement of cultures and languages. This is now possible with archaeogenetic evidence using genetic analysis. These studies have already been started and may in the future clear up some misconceptions in these regards.

For further information about this topic, start with either of the two books: The Invasion That Never Was by Michael Danino or In Search Of The Cradle Of Civilization by Feuerstein, Kak and Frawley. Their thought-provoking books will refer you to the available source materials on the subject. For a presentation by a top scholar in the archeological field on this topic see The Saraswati Flows On: the Continuity of Indian Culture by B.B. Lal.

Vedic Sanskrit versus Classical Sanskrit

Within the Vedic literature that has survived can be noticed subtle changes in grammatical forms and their usage from the earliest hymns to the concluding Upanishads (Vedanta). Various dialects and grammars of the language flourished in different regions and at different times within this Vedic period. Subsequent to the Vedic literature and Vyasa, grammar became formalized under Panini who lived at least two thousand years ago, and possibly even much earlier. Panini wrote a definitive grammar called the Panini Sutras, the oldest surviving grammatical work in the world. This work was so complete and influential that prior Sanskrit grammars, of which no fewer than sixty-four of their authors were mentioned, disappeared. The language had thus become formalized and has remained, for the most part, grammatically unchanged to the present – although some scholars have argued that the Sanskrit literature since the Vedic period, and even since Panini, has over time “suffered” a decline in grammatical richness.

A few language scholars claim that this apparent grammatically frozen nature of Sanskrit means it became a dead language. This is synonymous to saying a youth who has reached puberty, after which there is no further significant changes, except decline into old age, is “dead.” If it can be argued that English, or any other national or international language, is currently losing grammatical precision or richness, does this mean that that language is now “dead?”

Others may try to resuscitate the claim by saying that what they mean by “dead” is that it is no longer a language “commonly” used. However this claim suffers in at least two ways. The first is that they can offer no one number of speakers of a language that could reasonably quantify what “commonly” means, which would disqualify Sanskrit now as a “living language,” yet retain many other accepted living languages of the present or of the past spoken in certain areas or by certain tribes of people all over the world throughout history. Secondly, if “commonly” means predominately used throughout the daily life of its speakers, or, in other words, it is living when people choose to converse and formulate ideas in it in preference to any other, then they would have to prove that Sanskrit was ever the predominate daily language of its speakers, after which it became non-predominate and hence “dead” according to them. But what scholars of the language know of its history is that is was from earliest times always described as one of several languages known to its speakers, just like in present day India where most people are multi-lingual, changing the language to suit the listener or the situation.

Sanskrit was employed in several ways, such as for technical treatises, including scriptural treatises, where precision of expression is of utmost importance, for the language of the royal court, for the many daily rituals that occupied a substantial amount of the peoples time, and for a common language among a diverse linguistic audience. There are many different types of languages that are not the predominant language of anyone, such as legal language, various computer languages, mathematical languages, signal, code and sign languages, etc. These specialized languages we may call “dead” once they are no longer employed or known by anyone, but such is not the case with Sanskrit, of which quite a large number of people are currently skilled to varying degrees. I would venture to guess that more people are currently as skilled in Sanskrit, as those who are equally skilled in most any one of these other specialized languages, even though they are all simpler to learn than Sanskrit. In other words, “dead” just does not appear to be an objective, valid description of the language you are attempting to learn today. Now if “dead” is a valuation term meaning “useless”, then the fact that you are interested in this book shows, at least to your own satisfaction, that Sanskrit is not a dead language.

Whatever the status or stage of the language now, most call the earlier form of the Sanskrit language found in the Vedas as “Vedic Sanskrit” and the form of the language that conforms to Paninian grammar as “classical Sanskrit.” In this book, the grammar described is of the later, referred herein as simply Sanskrit.

If Panini wrote the definitive grammar book, then why are we offering this work?

Panini’s Sutras are in Sanskrit, so, if you don’t know the language and you wish to learn it, the Sutras themselves can’t help. Additionally, Panini himself addressed his grammatical work to those who were already familiar with the Sanskrit language and even speakers of that language, much like grammar is taught in all modern languages today to children who already are speakers of the language. Moreover, the Sutras are extremely brief, like mathematical expressions, requiring extensive enfoldment by a scholar of the language. Existing scholarly commentaries are available, but they are also in Sanskrit. So eventually you are driven to finding a teacher for these Sutras. There are many available, but being totally dependent upon the teacher for this study, you must commit many years of regular study with that teacher. Moreover, the Sutras themselves are extremely analytical and exhaustive on each topic throughout, so you cannot cut short this period of study and still get a broad enough grammatical basis to understand the language.

To address these issues many ancillary works are employed by teachers of the Sutras. These include a commentary on the Sutras, a book on verbal roots and their paradigm formations in verbal usage, a book on words and their paradigm formations into nominal and adjectival usages, a book of synonyms for basic vocabulary, a Sanskrit reader or examples from literature, sometimes (lately) a general Sanskrit dictionary, and often a simplified primer or handouts on some basic grammar topics. The problem with this massive collection is that a tremendous memory is required to retain and meld it, their vocabularies don’t match and the separate books don’t cross-reference or complement each other. Finally, this total package is best designed for a usage of the language that few beginning students will ever employ, namely to be full users of the language – readers, writers and speakers. Whereas, the vast majority of these students will only be readers of the Sanskrit language – quite a simpler skill.

Colleges in the West, and high schools and colleges in India, where a more limited length of study is available, solve this by employing Western style grammar books and usually a Sanskrit reader or examples from literature, along with the obligatory dictionary. Here the problem is not so much the method, but the materials provided.

The Sanskrit to English dictionaries for students that are chosen most in India are poorly printed editions that makes their use very frustrating. The readable dictionaries are expensive, and most are too complicated for the beginning student. The later also contain a mixture of Devanagari and English transliterated Sanskrit text that lowers their usability and hinders memory retention of what was just looked up.

The Sanskrit readers or examples from literature have little connection with the vocabulary, contents, or layout of the grammar book. Grammar books are of two styles – lesson style and reference style. Some lesson style grammars don’t complete the full spectrum of grammar necessary for a reading proficiency to go beyond their simple lesson sentences. Others provide too many grammar rules for beginning students, such as giving a grammar rule for isolated, individual forms that the student is never likely to encounter and would be better off just having the finished form of the word as a vocabulary entry. In other words, they admittedly are meant to prep the student to pass some final exam, rather than providing him or her with just enough to enjoy reading Sanskrit. Their vocabulary and practice sentences are sometimes offensive to Indian culture, reflecting the ruling British attitude of their day, or are geared towards children, or are collections of statements out of context that make them too difficult to understand and may even distort the culture as it sees itself. They therefore have little application to the literature a mature adult would want to read today. These lessons also break up the layout of the grammar and force the presentation order of the grammar to fit the lesson sentences, so that the grammar itself is disjointed and not nearly as effective as a unified, logical presentation would be.

Most of these grammar texts are not well printed, making them difficult to read and unnecessarily frustrating to the beginning student. The reference style grammar texts are mostly around 100 years old. They employ a Devanagari script, and/or an English transliterated script, of Sanskrit that contains a handful of archaic characters that are no longer used in printed texts. Several employ so much transliterated text – pandering too much to their students more interested in linguistic theories – that the student who simply wants to be able to read Sanskrit texts cannot relate it well to the literature in the Devanagari script with which he or she is faced. Practically all of them try to mimic or recreate every rule and exemption to the rule given in the major commentaries to Panini.

The best of the reference style grammars is Arthur Macdonell’s A Sanskrit Grammar for Students, a very well planned, logical, and especially concise presentation. I have based the following grammar on its best elements.

Lately there have cropped up teach yourself books. A perusal of them clearly shows that they present way too little grammar, and the grammar they present is usually so drowned in English prose explanations that the point of each topic is forgotten as the student turns the pages. Those books that don’t overwhelm the student with explanations fail in not explaining enough, or even any, of the grammar. They also employ too much of their Sanskrit text in English transliterated script, instead of the native Devanagari script, leaving the student ill prepared to even begin reading any useful or interesting Sanskrit text. There is one Indian teach yourself booklet series that uses the Devanagari script throughout by Dr. Pandit S. D. Satwalekar called Sanskrit Self Teacher. It avoids most problems of its genre, but its grammar presentation suffers many of the problems of the other lesson-style grammar books. I have based The Aruna Coursebook on its best elements.

Why is this series better for learning Sanskrit? And to whom is this work intended?

Why it is better is best addressed by describing the intended student. Firstly, if you are an Indian with a typical good memory for language, have an initial Sanskrit vocabulary provided by your native language and culture, and are young and have many years for study under a teacher, or are older and have the leisure time to study under a teacher for even more years, then use the Paninian method. This method has over thousands of years of perfection and can take you as far as you want to go in Sanskrit. However, if you are not one of the above or are one of the above and are older, you just might want to employ this work first. It will quickly get you to where you can experience the enjoyment of reading Sanskrit before undergoing the admitted stress of in-depth Paninian study. And it will get you to the proficiency in reading Sanskrit that Panini would have expected from his beginning students.

More important to many of these types of students, as well as to others, will be the accomplishment of knowing in great depth the Bhagavad Gita, the finest teaching on reality available anywhere in any language. The course should take from one to four years of daily attendance, depending on your starting background and the quality of effort put forth. This is a serious, academic work, even though in a self-teacher format, meant to provide the equivalent of two years of college-level Sanskrit study. You are encouraged to study two to three hours daily – short of this, it will take you longer to get through the material. The first one to two hours is (eventually) stress-free, daily reading of the Bhagavad Gita text. This practice has the added benefit of giving you a tremendous familiarity, if not memorization of the text. At the completion of the course you will know the Gita in the original better than any translation ever made, and that knowledge of Sanskrit and the Gita will be where it should be – in your head, instead of still in a book.

The intended audience for these books ranges from the very beginning Sanskrit student to the advanced student wanting technical details and explanations on the Bhagavad Gita, and includes the non-Sanskritist who is willing to wade through a little Sanskrit in order to clearly know the Bhagavad Gita and its mature Vedic vision.

The Aruna Sanskrit Language Series

The first title in The Aruna Sanskrit Language Series is The Aruna Sanskrit Grammar Reference. The Grammar Reference follows the reference style presentation of grammar. This will provide a logical, consistent and complete exposition of the grammar for reading-only proficiency. The grammar uses minimal wording and presents the material in outline and chart form as much as possible to maximize your visual memory of the information. This method also greatly assists review and re-review of the grammar, necessary to master the subject. Several of the charts have multiple forms within them that demonstrate simple, easy to remember patterns which are overlooked by many grammar books. This helps you to also logically remember the information. The grammar is presented in the Western style, which emphasizes analytic, reading skills, as opposed to the more difficult composition and speaking skills. As much as possible, the examples in the book are taken from the Bhagavad Gita, so your targeted vocabulary builds quickly.

The second title in The Aruna Sanskrit Language Series is The Aruna Sanskrit Grammar Coursebook: 64 Lessons Based on the Bhagavad Gita Chapter Two. The exercise in the first lesson of the Aruna Coursebook directs the student to the Script Reading Exercise, given in its appendix. This section provides the complete Bhagavad Gita second chapter in large-print Sanskrit with transliteration using the English alphabet under each line, followed with the entire chapter again with only the large-print Sanskrit to test your progress. The text and transliteration are broken down in two separate ways to show the separate syllables and then the individual words, thus progressively showing the student the proper methodology for correctly pronouncing the original Sanskrit text. This section should provide all the necessary practice material for the student to learn the Sanskrit script – essential for proceeding through the rest of this work and any other Sanskrit work. For students who need help in pronunciation of Sanskrit words, I highly advise finding a teacher, a friend or someone in your community who will surprise you with their readiness to assist you – knowledge of Sanskrit and its literature seems to nurture this helpful attitude. These people need not know the meaning of all the Sanskrit words, but they can read the Sanskrit script. Additionally, a tape or CD of the Bhagavad Gita is available through the Arsha Vidya Gurukulam (

An alternative help for this Script Reading Exercise is the specially developed Sanskrit Reading Tutor: Read It, Click It, Hear It!, a uniquely formatted PDF file that has the alphabet sections from the Grammar Reference plus the Script Reading Exercise of the Grammar Coursebook. The special feature of the PDF is that one can click on any of the characters in the alphabet section to hear its pronunciation, and on any of the individual lines, quarter verses, or their syllables of Script Reading Exercise to hear their pronunciation. In this way one can have each of these script elements individually re-read to you over and over while you are seeing the script on your screen – a unique and invaluable tool for learning the script and its pronunciation.

The rest of the lessons in the Aruna Coursebook give a relatively quick overview of the entire Grammar Reference – its layout, its methodology and how its grammar rules are to be understood. Enough explanation with little redundancy is provided to explain the outlined and charted grammar presented in the Grammar Reference so that you can learn on your own with little or no assistance from a teacher, assuming you can apply a post-high school aptitude towards this work.

The examples and exercises in the Aruna Coursebook are drawn from the second chapter of the Bhagavad Gita, which is taught completely in prose format, then finally in verse format. In the Aruna Coursebook, vocabulary is presented by giving the English meaning of the individual words, their grammar and also their contextual use within expressions from the Gita. This should make it very easy to learn the vocabulary. By teaching the vocabulary of the verses in a consistent Sanskrit prose order, which is syntactically based, the student gains a quicker grasp of Sanskrit syntax. All exercise expressions and sentences are taken directly from the Gita – no extraneous material is presented. These words, expressions and sentences are repeated throughout the Aruna Coursebook in the lessons and exercises to assist your memory; nothing is taken for granted as you progress through the Aruna Coursebook. All the examples and lessons are cross-referenced to their specific verses in the Gita, so the context and the English translation given therein will provide an answer-key and guide for the student’s work.

After knowing the prose order of all the verses of the second chapter, the student is then introduced to the analysis of verses so that he or she can quickly and confidently see the meaning of the Bhagavad Gita verses directly. Concepts presented in the Gita are expounded, so that non-Indian and Indian students alike can appreciate the depth of discussion within the Gita.

The Aruna Coursebook forms a four to twelve month full-time course in itself, depending on the effort put forth and any previous exposure to this language or familiarity with other languages that may have affinity to the Sanskrit language. At the completion of The Aruna Coursebook you will be familiar with the entire range of Sanskrit grammar for reading proficiency and be able to sight read with understanding all the verses of the second chapter of the Bhagavad Gita, which briefly presents the entire teaching of the Gita.

The third title in The Aruna Sanskrit Language Series is The Bhagavad Gita Dictionary. This is a Sanskrit to Sanskrit and English dictionary of the entire Bhagavad Gita. Many entries include reference citations to grammatical rules for their peculiar formations presented in Grammar Reference. The derivation of many words is included as appropriate. The entries also provide Sanskrit synonyms and/or meanings, plus contextual Sanskrit analysis of compound words, so that your vocabulary builds beyond the Gita with each use. The goal of this vocabulary building is to encourage you to start to think in Sanskrit while you read Sanskrit – a multiplying effect that greatly enhances learning and builds the skills necessary to progress to Panini and the full grasp of the Sanskrit language, if you so choose.

The fourth title in The Aruna Sanskrit Language Series is The Bhagavad Gita Reader: Sanskrit/English Parallel Text, consisting of all eighteen chapters of the Bhagavad Gita. It presents on each left-hand page the Gita verses, then on the opposing page the same verses in easier to understand prose order and an English translation of those verses, in columnar format.

The purpose of this section and its design is three-fold. First, you are instructed to read Gita verses daily. Reading skills are encouraged and developed so that you can read Sanskrit text as effortlessly as your native language. By reading down each of the respective pages, the student can quickly read either the original verses, the prose version, or the English rendering – as well as relate these three by reading across the opposing pages. Later understanding of what you are reading follows and is greatly enhanced by this exercise.

Secondly, the prose presentation of the verses quickly brings you to an understanding of the meaning of the verses, well before the difficult skill of unraveling the grammar packed into verse form. Breaking with tradition, all words in the prose are grammatically split apart, helping you see the individual words with their full grammatical form. Interspersed in the prose in parentheses are additional Sanskrit words necessary to help you comprehend the meaning and context of certain words, and to understand the not-so-obvious references of certain pronouns found in some verses.

Thirdly, the English translation is given out-of-the-way in its own column to de-emphasize your dependence on English as a medium for understanding Sanskrit. This translation doubles as a quick answer-key to the Aruna Coursebook exercises, which are all cross-referenced to the matching verses. Additionally, contextual explanation is added in parentheses so as not to mislead the reader into confusing editorial commentary with the actual translation. This is a feature sorely lacking in existing translations that I have come across. Certain Sanskrit words that have a depth of technical and cultural meaning packed into them, or intentionally have multiple meanings, are well explained in the Aruna Coursebook and in The Bhagavad Gita Dictionary. After being initially translated, the original Sanskrit word is then used in the English translation of the following verses and is clarified, if necessary, only in parenthetical commentary. This makes for better readability of the translation, during this language study and especially after this study – once you realize that there are no equivalent, concise expressions in English for certain Sanskrit words, nor need they be manufactured.

The fifth title in The Aruna Sanskrit Language Series is The Bhagavad Gita Sanskrit Key: Verse-by-Verse Grammar & Vocabulary. The Gita Key also can be used as a quick answer-key to the Aruna Coursebook exercises. By collecting the vocabularies together, this becomes a marvelous tool to study the grammar, syntax and meaning of the verses of the entire Bhagavad Gita.

The Aruna Sanskrit Language Series was arranged for the following reasons. The Grammar Reference and the Aruna Coursebook are meant to stand together as an introduction to Sanskrit, without needing the other three books. If the student wants to proceed in this learning of the Bhagavad Gita, then minimally the Gita Key is also required. The Gita Dictionary was separated from the other books as it is more convenient to have a separate dictionary that can be opened as a reference for the Aruna Coursebook, Gita Reader, and Gita Key, or as a reference tool apart from this series. If one is not interested in learning Sanskrit grammar and can read the Sanskrit script, yet wants to know the Bhagavad Gita and how its meaning is arrived, then just the Gita Key will more or less suffice, though its thousands of grammar rule references require the Grammar Reference. If one simply wants a very good translation of the Bhagavad Gita and can read the Sanskrit script, then the Gita Reader is complete in itself. If you cannot read the Sanskrit script, but still want to see this same good translation, then see this author’s The Bhagavad Gita: Victory Over Grief and Death. It has the same English translation of the entire Bhagavad Gita within it, with the number of Sanskrit words, in transliteration, retained in the translations greatly reduced from the Gita Reader version.

Although not specifically designed as a classroom text, the various components of this work may be used as an aid or auxiliary to classroom instruction. On the other hand, this work specifically helps those who have students who are interested in learning Sanskrit, but who do not have the time, materials or expertise to teach the subject themselves, yet can act as an occasional or regular coach for these students.

If the reader is already familiar with or knowledgeable in Sanskrit, then this series provides the benefits of a quick but thorough reference guide to the grammar of written Sanskrit for personal or teaching purposes, and an in-depth analysis of each word, each compound-word, each sentence, and each topic of the Bhagavad Gita – in accordance with the ancient core tradition, free from any new-age accretion. Therefore nothing stands between you and the original text. You will understand in English, what Arjuna heard in Sanskrit from Lord Krishna.

Go to Top ('Home' key on many Windows browsers)

1.1–4 1.5 1.6–7 1.8–9 1.10–1 1.12–6 1.17 1.18–9 1.20–4 1.25

The Sounds in the Sanskrit Language

1.1: The script. Saṃskṛta is the actual name of the language, but commonly is called by the Anglicized Hindi word Sanskrit. Today, Saṃskṛta is generally written in the Devanāgarī script, although some texts employ a regional script; for example, a version of the Tamil script in the southern state of Tamil Nadu. The Devanāgarī script itself has varied over time (even within the past century). The character formations used in this book are the current standard. The Devanāgarī alphabet consists of 48 characters, for the basic 13 vowel and 35 consonant sounds.

1.2: The sounds. The ancient grammarian Pāṇani recognized all 48 of the Saṃskṛta sounds and scientifically grouped them under a number of classifications, depending on the purpose. One classification (listed in 1.3: using Western terminology) is the location of the sound as it is made by the human voice. Another classification is the effort of the contact within the mouth at that location (1.4:).

1.3: The five locations of sound.
Most characters have one location; some blend two locations; two ( and ) vary, depending on the sound that precedes them; and one of those two () varies, depending on the sound that follows.
GutturalFormed by contact of the base of the tongue against the back of the throat (kaṇṭha).
PalatalFormed by contact of the middle of the tongue against the palate (tālu). The tip of the tongue touches near the front palatal ridge (the top of roots of upper front teeth, a half-inch above the top of the teeth).
CerebralFormed by contact of the tip of the tongue against the front of the downward hard dome (the mūrdha) at the middle of the palate. There is no good equivalent of this sound in English.
DentalFormed by contact of the tip of the tongue against the base or root of the upper front teeth (danta).
LabialFormed by contact of the lips (oṣṭha).

1.4: Five efforts of articulation and the characters that exhibit them.
The efforts are the extent of contact, or the shape of the restriction, between the organs of articulation where the sound is generated.
TouchedThe five classes of consonants (k – m in the alphabetical table, 1.5:).
Slightly TouchedThe semivowels (y – v).
Slightly OpenThe sibilants (ś – s), h, visarga (), and anusvāra () (see 1.5: ).
OpenThe vowels except a (ā - au).
ContractedThe short vowel a.

1.5: Devanāgarī alphabet with international transliteration, sound, and location. (Alphabetical order).
Initial MedialTransit.Sounds LikeLocation
  aao in songuttural
◌ाāo in bottle
िie in bepalatal
◌ीīe in bee
◌ुuo in movelabial
◌ूūoo in moon
◌ृrh in rhythmcerebral
◌ॄin rhrhythm
◌ॢle in tabledental
◌ेea in tapegutteral-palatal
◌ैaiy in my
◌ोooe in toegutteral-labial
◌ौauow in now
Initial or MedialTransit.Sounds LikeLocation
 bhalf an hguttural or labial
 cn in French bonconforms to preceding vowel
kck in blockguttural
khckh in blockhead
gg in log
ghgh in log-hut
ng in song
cch in catchpalatal
chchh in catch him
jge in hedge
jhgeh in hedgehog
ñn in cringe
t in hurtcerebral (1.3:)
ṭhth in hurt him
d in gored
ḍhdh in gored him
n in corn
tt in catdental
thth in cat hair
dd in mad
dhdh in mad house
nn in numb
pp in looplabial
phph in loop-hole
bb in rob
bhbh in rob him
mm in much
yy in youngpalatal
rr in dramacerebral
 dll in luckdental
v (w)v in vine /
w in Swami e
labial-dental / bilabial
śsh in shippalatal
sh in perishcerebral
ss in sitdental
 fhsoft h in humguttural
There is a tendency to slightly aspirate initial mutes, such as “k,” “t,” and “p” in English. In Saṃskṛta, initial nonaspirate mutes (1.9:) are more like these sounds at the end of isolated English words – where the aspiration is mostly cut off.
a.Noninitial short अ a has no character, as it is inherent in every consonant from क ka to ह ha. Consonants without any vowel after them are marked below with a stroke slanting left to right ् (called a virāma or halanta), as in क् k. When so written, the consonant is referred to by the term –kāra. For example, क् k is called ककार kakāra and ख् kh is खकार khakāra.
Consonants by themselves without either a following or preceding vowel are considered not pronounceable. (We unknowingly admit the same in English. The letter b is pronounced “be,” f as “ef,” etc.) To pronounce each Saṃskṛta consonant in table 1.5:, a final अ a is added (e.g., क is read “ka”, but the description and classification there applies only to क् kakāra).
Sometimes when a single consonant (e.g., a mute or न् n 1.9:) is at the end of a word with a pause afterward (e.g., at the end of a sentence), we add a short echo of the preceding vowel. For example, in तत् tat, the अ a is echoed: tata.
b.Visarga (ः ) never occurs as a word initial. Like other consonants, visarga cannot be pronounced by itself without a preceding vowel; moreover, visarga cannot be joined in writing with a following vowel. Visarga corresponds to the second half (-h) of the hard aspirates ख् k-h, छ् c-h, थ् t-h, and फ् p-h (1.9:). It is usually pronounced from the same position as its preceding vowel – as a short, hard blowing out of the breath.
Before the gutturals क् k and ख् kh, the visarga is a guttural – a distinct, hard blowing with constriction at the base of the tongue – and is called jihvāmūlīya. Before the labials प् p and फ् ph, the visarga is a labial and has a distinct bilabial “f” sound called upadhmānīya.
When at the end of a word with a pause afterward (e.g., at the end of a sentence), visarga is pronounced as a hard “h” followed by a short echo of the preceding vowel. For example, कृष्णः kṛṣṇaḥ is pronounced kṛṣṇaḥa. When vowel is ऐ ai or औ au, the echo is of the component इ i or उ u (2.3.a:) respectively, e.g. नौः nauḥ (3.41:) is pronounced nauḥu.
The optional visarga (before a sibilant, 2.28:) alphabetically occupies the place of the sibilant it replaces and is pronounced as the sibilant. For example, in अन्तःस्थ antaḥstha, the visarga has replaced the sibilant स् s that was in the original अन्तस्स्थ antasstha.
c.Anusvāra (ं ) never occurs as a word initial. Its pronunciation is like the French nasalization of their vowels. For example, the French word bon in bon voyage would be written and pronounced in Devanāgarī as बां bāṃ.
The optional anusvāra (before mutes and the nasals न् n and म् m, 2.55:) alphabetically occupies the place of the class nasal it replaces. For example, compare शंकर śaṃkara and शङ्कर śaṅkara (where anusvāra replaces ङ् ) or – incorrectly (1.10:) – कंपित kaṃpita and कम्पित kampita (where anusvāra replaces the root medial nasal म् m).
d.ल् l often interchanges with, or is derived from, र् r.
e.व् v is usually pronounced as a labial-dental “v,” made with the lower lip first slightly touching the upper front teeth. When preceded by a consonant in the same syllable, however, it is pronounced as a bilabial “w.” This has led to the popular transliteration of the conjunct व् v as w in many Saṃskṛta words that were coined in another language. For example, स्वामी should be pronounced and is popularly transliterated as Swāmī, instead of Svāmī, although the latter is technically correct.
f.ह् h corresponds to and is derived from the second half (-h) of the soft aspirates घ् g-h, झ् j-h, ध् d-h, and भ् b-h (1.9:). It is pronounced from the same position as its following vowel.

1.6: Writing. When writing Devanāgarī, the distinctive part of each letter, occupying at least two-thirds of the height of the final letter, is normally written first, then the vertical line, and finally the top horizontal line. For example, the syllable त ta is written with three strokes, the last two being the vertical line then the horizontal line. But, write in whatever order works for you. The top horizontal line is usually drawn after the rest of the word or phrase has been written. In fact, it was regularly missing in old manuscripts.

1.7: Writing noninitial vowels. Each vowel is written in a different way, according to whether or not it is initial. There is no sign for the noninitial short अ a (as noted in 1.5.a:). The noninitial short इ i is written before the consonant after which it is pronounced (e.g., कि ki). The rest of the noninitial vowels are written after, below, or above the consonants. An example is the letter क् k with all its vowel signs: क ka, का , कि ki, की , कु ku, कू , कृ kṛ, कॄ kṝ, कॢ kḷ, के ke, कै kai, को ko, and कौ kau. Certain consonants take certain vowel signs at the side instead of below. They are रु ru, रू , and हृ hṛ. When the consonant र् r precedes the vowel ऋ , it is written as a hook on top of ऋ (e.g., as र्ऋ rṛ) and is pronounced before the vowel. Note that the vowel ऋ is written as an initial vowel even though it is medial (e.g., निर्ऋतिः nirṛtiḥ). The consonant श् ś is often written in its alternant form when followed by the vowel ऋ or ॠ (e.g., शृ śṛ) and sometimes also when followed by the vowel उ u or ऊ ū.

1.8: Sound classifications. The 48 characters of the Devanāgarī alphabet are classified by sound as guttural, palatal, cerebral, dental, or labial (see table 1.9:). They are further grouped as mute/nonmute, hard/soft, class/nonclass, consonant/vowel, aspirate/nonaspirate, nasal/semivowel/sibilant, simple/diphthong, and short/long (see Glossary). The consonants have an added अ a for pronunciation purposes only (1.5.a:). Committing table 1.9: to memory is helpful for understanding sandhi (see chapter 2). Traditionally the consonants (and vowels) in table 1.9: are read alphabetically: Read the rows of the first five columns across (क ka, ख kha, ग ga, घ gha, ङ ṅa, च ca, छ cha…). Read the next three columns down (य ya, र ra… ह ha). If in doubt, refer to table 1.5: for correct order.

1.9: Devanāgarī alphabet classified by sound.
ClassMute aNasal-
Class ConsonantNonclass ConsonantVowel
Gutteral  c d
Labial b  c
a.A mute is a sound that begins with a complete stop of the passage of the breath (1.4: touched), also called a stop in phonetics. Nasals are not full mutes, but are oral mutes with a nasal continuant (Macdonell 1927), thus in the following rules the term “mute” will not include nasals unless indicated.
b. va is labial-dental (1.5.e:).
c.Visarga (ः ) may appear at the end of the last word in a sentence, before a sibilant, or before a hard guttural or labial. After last word ः is pronounced in the location of the preceding vowel, and before sibilant as the same sibilant (see 2.28:). In the last two cases ः is pronounced in the location of the following guttural or labial (see 1.5.b:). Hereafter, visarga will be indicated with a colon (:) for better visibility on these pages.
d.Diphthongse and ऐ ai are guttural-palatal. Diphthongs ओ o and औ au are guttural-labial, per their component sounds (a-i for ए e and ऐ ai, and a-u for ओ o and औ au – see 2.3: and 2.3.a:).

1.10: Anusvāra (ं ), or “after-sound,” is a nonmute, unmodified nasal that follows a vowel. It is sounded only through the nose, not the mouth. For this reason it is not classified in table 1.9:. Anusvāra is sometimes written as ँ m̐, where it is then described as anunāsika (pronounced through both the nose and mouth). There can be a nasalized ल् l, written as ल्ँ (2.51:). Anusvāra is properly found before the semivowels, sibilants, and ह् h (2.54:), and optionally before mutes or nasals (2.55:). But it is sometimes incorrectly used within a word (i.e., not due to a general sandhi, as in 2.55:) in place of any of the five class-consonant nasals when they are followed by a mute or nasal, or at the end of a word or sentence (e.g., the incorrect अंगं aṃgaṃ for अङ्गम् aṅgam). This usage does not affect the pronunciation of the correct corresponding class nasal. For more about anusvāra, see 1.5.c:.

1.11: Common symbols. Saṃskṛta commonly uses these punctuation or stylized symbols.

DaṇḍaThe single daṇḍa । is used at the end of a line of verse and at the end of a sentence. The double daṇḍa ॥ is used at the end of a verse and (sometimes) at the end of a paragraph.
AvagrahaOptionally marks the dropping of a short अ a at the beginning of a word due to a sandhi (phonetic combination) with a previous word (2.12:). If the अ is nasalized, the anusvāra normally appears before the ऽ avagraha. ऽ is usually transliterated as an apostrophe (e.g., तेजोंऽश tejoṃ'śa for तेजो अंश tejo aṃśa).
AbbreviationMarks the dropping of a contextually understood part of a word. For example, गतेन gatena is abbreviated ॰तेन -tena in a dictionary entry, where ग- (ga-) is understood.
OmA stylized form of ओम् Om, a name for brahman (reality).

1.12: Conjunct consonants. If a consonant is followed immediately by one or more consonants – even when the consonants are divided between two words in a sentence – they are normally joined, after sandhi (chapter 2), into a cluster called a conjunct consonant. For example, त् स् t s are combined into त्स् ts.
The general principle in forming a conjunct consonant is to drop the right side vertical lines except in the last letter (e.g., त् t + स् s + य ya is त्स्य tsya). When the first letter has no right-side vertical line, or when both letters have a right-side vertical line and the following consonant has room for the previous letter to anchor to its vertical line (above the distinctive portion of its character), then – in both cases – the letters are conjoined vertically, with the initial consonant on top. For example, अङ् aṅ + ग ga becomes अङ्ग aṅga and अञ् + च ca becomes अञ्च añca. There are also irregular formations (see tables 1.13:).

1.13: Irregular forms in conjuncts. Certain characters morph their shape to fit the shape of the conjoined character.
क्morphs with (e.g.)क्+तक्तk·ta
श्श्+चश्चś·ca (see 1.7:)
The character र् r is called रेफ repha, instead of rakāra – 1.5.a:.
र्morphs with (e.g.)क्+क्रk·ra, when र् has a vertical line to attach to.
ड्+ड्रḍ·ra, when र् has no vertical line to attach to.
र्+तर्तr·ta, when र् precedes a consonant (or the vowel ऋ ), the mark is placed above the (conjunct) consonant and to the right of any following vowel marker (e.g., नैष्कर्म्ये naiṣkarmye, निर्ऋण nirṛṇa).

1.14: Special conjuncts. There are two conjunct consonants whose component letters within Devanāgarī are indistinguishable: क्ष k·ṣa, and ज्ञ j·ña. One common, current pronunciation of ज्ञ j·ña, which is preferred by this author, has the component ज् j sound very slightly heard. For this sound, place the tongue and mouth in the palatal position to pronounce the ज् j, then pronounce the palatal ञ् ñ for the length of two consonants (ञ्ञ् ññ). In some parts of India, another pronunciation is with the “j” component pronounced as the guttural sound “g”. Since the following nasal would tend to naturally sound as a guttural, then the conjunct ज्ञ is pronounced as if ग्ङ gṅa. This latter pronunciation probably came in from a regional dialect, but you will hear it.
In Vedic grammar and phonetics, ज्ञ falls under the Vedic rule that when a consonant is in conjunction with र् or a nasal, the Vedic meter demands that a very short vowel (a fraction of the length of a normal vowel, called a svarabhakti, Vedic Grammar Macdonell 1966) be pronounced between them. Hence, ज्ञ would have been pronounced as jaña, so it would seem that the initial ज् j would have had its proper palatal pronunciation. Other than this conjunct, unlike English, classical Saṃskṛta is normally pronounced as it reads.

1.15: Printing consonants. In print, Devanāgarī conjunct consonants are presented in different ways. The differences are sometimes due to the inability of a particular Devanāgarī font to completely form a conjunct consonant. Other times they are due to an editorial preference. An editor may want to avoid using long or visually difficult conjunct consonants that challenge the reader. In these cases, one or more of the characters of the conjunct, usually the first character(s), are written with a virāma (1.5.a:). For example, बुद् बुद bud·buda, instead of बुद्बुद budbuda.
It is common, though, to avoid using the virāma in print wherever possible, even by joining a consonant-ending word to a following vowel-initial word. For example, the three words अशोच्यान् अन्वशोचः त्वम् aśocyān anvaśocaḥ tvam would be written as अन्वशोचानन्वशोचस्त्वम् aśocyānanvaśocastvam, where न् अ n a join to become न na, and visarga before त्वम् tvam changes to स् s (see 2.27:) and joins with the following consonants into स्त्वम् stvam. Saṃskṛta was written on materials such as palm leaves, so the writing of the script needed to be compacted to fit the compactness of the media. Hence this natural avoidance of virāmas which leave gaps between words.

1.16: Reading conjunct consonants. To read a conjunct consonant, read the component characters left to right and, within that, from top down (e.g., द्वन्द्व d·va·n·d·va). The only exception is र् r at the beginning of a conjunct consonant (e.g., कर्त्स्न्ये kar·t·s·n·ye, in this example, र् r is written after the final vowel e but pronounced before the t – see table 1.13:).

1.17: Common conjunct consonants (in alphabetical order). As new principles in conjoining are introduced alphabetically in the list, their examples are dark-shaded. Many conjuncts can be written in multiple ways, such as stacked or side-by-side, e.g., क्च or क्च k·ca. The below conjuncts are according to the Siddhanta font, unless your browser lacks or overrides this font. Another Devanāgarī font may differ in which conjuncts employ which of these principles. A conjunction of two components in transliteration is indicated by either a hyphen (-), or in this chart by the more compact middle dot (·).

1.18: Numerals.
a.Saṃskṛta is the original source for numerical writing in the West; therefore the number 2002 is familiarly written २००२.

1.19: Numerals as pronunciation indicators. The figure २ after a word indicates that the word is repeated (e.g., अहो २ indicates अहो अहो aho aho). The figure ३ after a vowel indicates the pluta, or protracted lengthening, of the vowel to three mātrās.
The pronunciation length of time (mātrā) of a short vowel, such as अ a, is one mātrā. A long vowel, such as आ ā, or a diphthong is two mātrās (twice as long as अ a). आ ३ indicates the vowel is three mātrās. The three-mātrā pronunciation is found only in the Veda, and there it is infrequent. The pronunciation length of one consonant is said to be a half mātrā.

1.20: The syllable. Like in English, a phonological (or pronunciation) syllable, is called an अक्षर akṣara (also called a मात्रा mātrā, but not to be confused with the pronunciation length of time in 1.19:). It is the smallest unit that can have a meaning, and is the building block of words.

1.21: Syllable structure. A syllable is centered on a vowel.
With/without a preceding consonant or conjunct consonant+Vowel+With/without a following consonant or two (including anusvāra or visarga), to complete a meaningful component of a word.
E.g., अशोच्याननन्वशोचस्त्वं consists of eight syllables: a-śoc-yān anv-a-śoc-as tvaṃ.

1.22: Quality of syllable. A syllable is heavy (गुरु guru) if its vowel is long or a diphthong, or if its vowel is followed by an anusvāra, visarga, or conjunct consonant (even if the conjunct consonant is in the next word or the result of a sandhi with the next word within the same pāda 1.23:). Otherwise, the syllable is light (लघु laghu). For example, the first syllables in the words bāl-a, bodh-i, duḥ-kha, and bud-dhi are heavy, and the second syllables are light. The quality of syllables may determine how a grammatical rule is applied to form a word stem. It also helps determine the meter (1.24:) of a verse.

1.23: The pāda. Much of Saṃskṛta literature is in verse form, and most verses are in the form of stanzas of four metrical quarters. A quarter is called a पाद pāda (literally, “foot”). Two pādas make a metrical line. The end of a metrical line is treated in pronunciation and writing like an end of a sentence. In accentuation (1.25:) though, each pāda is treated like an independent sentence; for instance, an enclitic (see Glossary) can never begin a pāda (4.3.1:).

1.24: Meter. The quality and the quantity of the syllables in a pāda determine the meter of the verse. The meter is the recurring pattern of the heavy and light syllables, forming a patterned rhythm. Anuṣṭubh, or Śloka, the most common meter in Saṃskṛta literature, is composed of two lines of two pādas, with eight syllables to a pāda. Occasionally three verses can be arranged into two triplets (three lines each). Each verse is typically a complete sentence. It is a particularly flexible meter, because of its wide variety of metrical forms. A metrical form defines how the light and heavy syllables are patterned in a line. In the most common metrical form of Anuṣṭubh by far, called Pathyā, the 5th-6th-7th syllables of the first pāda are light-heavy-heavy, and of the second pāda they are light-heavy-light, all the rest of the syllables of the line can be either light or heavy (see Macdonell 1975 Appendix II).

1.25: Accent. Saṃskṛta in Vedic literature has a tonal accent, in which the accent of a vowel is in the form of a change in pitch: raised (उदात्त udātta), lowered (अनुदात्त an-udātta), or transitional (स्वरित svarita – moving from a raised-tone semivowel y or v, indicating their original i and u vowel pronunciation, to the unaccented vowel that follows the y or v).
Later Saṃskṛta is pronounced with a stress accent instead of a tonal accent. In a stress accent, the accent of a vowel is pronounced more forcefully or loudly relative to nearby, unaccented vowels. Whereas there are rules and text markings for the placement and nature of the tonal accent, there are no specific rules or text markings for the stress accent. Although the placement of the stress accent in a word is generally deduced from the rules regarding the Vedic raised tonal accent, the stress accent may have shifted, depending on the quality (heavy or light) of the syllables.
In the case of the many new words introduced since Vedic literature, where the placement of the accent may not be clear from derivation, generally the accent falls on the last metrically heavy syllable within the word (e.g., Kalidā̍sa, Himā̍laya, kā̍rayati). The stress accent, like the Vedic tonal accent, may then disappear or move to another syllable, depending mostly on the word’s inflection (nominal declension – see Glossary and 3.3: – or verbal conjugation – see 5.3:), or placement in a sentence.
Like in any other language, accent in Saṃskṛta is best learned by listening to an experienced, native speaker, but a few guidelines are as follows:

  • The first finite verb in a main clause loses its accent, unless it begins the main clause or pāda. Subsequent verbs in the same main clause retain their accent, being treated in accent as if beginning their own sentence.
  • A word in vocative case is treated as having no syntactical part of a sentence (i.e., as if it is a separate sentence unto itself). As a result, a finite verb that follows the initial vocative(s) in a sentence or pāda is treated as if it were at their beginning. A vocative is always accented on its first syllable (3.7:), but the vocative loses the accent unless the word is at the beginning of the sentence or pāda.
  • For the treatment of unaccented, enclitic words, see 4.3:.
  • In certain sections of this text, for example, when a grammatical tonal accent rule is specified, the accent is marked (cf. Macdonell 1979). Note that a stress accent may shift if the syllable is not heavy. For example, the causal verb kāra̍yati, with its tonal accent on the causal suffix a̍ya (5.40:), may instead have its stress accent on the preceding heavy syllable kā̍rayati. For a discussion of Vedic tonal accent, see the appendix of A Vedic Grammar for Students (Macdonell 1966).

Now you have an understanding of the alphabet.

If you have The Aruna Sanskrit Grammar Coursebook: 64 Lessons Based on the Bhagavad Gita Chapter Two, it will help guide you through the study of the rest of this Grammar Reference. The Aruna Coursebook will direct you to the following chapters in this book as and when required.

Alternatively, if you do not have the Aruna Coursebook but need further excercise help to better read Devanāgarī text, you can order The Sanskrit Reading Tutor. This pdf file has audio clips attached to the text so you can read it, click it and hear it. The text in the pdf is the same as the Script Reading Exercise in the Aruna Coursebook appendix and covers the entire second chapter of the Bhagavad Gita.

Go to Top ('Home' key on many Windows browsers)

2.1 2.2–3 2.4 2.5 2.6–22
2.23–4 2.25–40 2.41–64 2.65–112

2.1: Coalescence of adjacent sounds. In Saṃskṛta every sentence is treated as a continuously spoken sequence of sounds, written exactly as they are pronounced. The coalescence of these sounds when pronounced as a sequence is called sandhi. The purpose of sandhi is to facilitate smooth pronunciation, and to avoid a gap between vowels in separate syllables, called a hiatus by Western grammarians.
Although several consecutive consonant sandhis may occur for easier pronunciation of a consonant cluster, only one vowel sandhi is applied between two adjoining vowels. Any hiatus in a finished Saṃskṛta text is either the result of only one vowel sandhi, in which case additional application of sandhi to avoid hiatus would render the original syllables unidentifiable, or the absence of any sandhi application at the end of a metrical line or the end of a sentence.
Because of sandhi, generally the ends and sometimes the beginning of words will be written in a sentence as they orally sound in combination, and not as they may be found separately, e.g., as in a dictionary.

a. The following sandhi rules, as well as many other grammar rules, are subject to occasional irregularities, of which only the common exceptions are noted herein. Not infrequent is the lack or irregular application of sandhi to maintain meter, such as retaining hiatus between two pādas (1.23:).

2.2: Guṇa and vṛddhi. Vowels may strengthen by taking guṇa or vṛddhi. The simple liquid vowels (i.e., इ ई, उ ऊ, ऋ ॠ including ऌ, theoretically), as well as these vowels as components of diphthongs (i.e., the non-simple liquid vowels), may also be changed into semivowels, or those semivowels may revert back to their liquid vowel, called samprasāraṇa.

2.3: Strengthening of vowels.
Simple vowelsGuṇa (components)VṛdddhiSemivowel a Samprasāraṇa (weakening) a
इ ई(=अ+इ/ई)(=अ+ए)य्
उ ऊ(=अ+उ/ऊ)(=अ+ओ)व्
ऋ ॠअर्(=अ+ऋ/ॠ)आर्(=अ+अर्)र्अर् or र
अल्(=अ+ऌ)(no vṛddhi for ऌ)
a.The simple liquid vowels (including the second half component इ or उ of diphthongs ए ऐ or ओ औ respectively) may become (or strengthen to) semi-vowels य्, व्, र् respectively. In reverse, these semivowels may weaken, called saṃprasāraṇa, to their respective simple liquid vowels. Notice that the semivowel’s following or preceding अ is dropped in saṃprasāraṇa.

2.4: General and internal sandhi. There are two forms of sandhi.

General sandhi between:
Independent word+ Independent word
Word in a cmpd.+ Next word in the cmpd. (6.29:), including verbal compounds (6.11:)
Nominal stemSecondary suffix beginning with any cons. except य् (6.27:)
Pada declension terms. - भ्याम्, भिस्, भ्यस्, सु (3.2.a:) – First apply general sandhi rules 2.23: & .24:, then internal sandhi rules will apply.

Internal sandhi between:
Other combinationsApplicable specific internal sandhi rules apply first, then general sandhi rules. With the exception that finals of roots usually don’t change, nor generally would rule 2.23: apply.

2.5: Reading sandhi charts. In the following sandhi charts, unless otherwise noted, the Preceding column’s letter(s) alone change.
For example, rule 2.9: should read – when अ आ (i.e., अ or आ) precedes a guṇa vowel, then both (अ and following guṇa vowel, or आ and following guṇa vowel) are changed to the following’s component liquid vowel’s vṛddhi. In this case, the component liquid vowel being इ or उ of the guṇa vowel ए (=अ+इ) or ओ (=अ+उ) per 2.3.a:, its vṛddhi would be ऐ or औ (2.3:), respectively.
Rule 2.14: should read - when guṇa vowel ओ precedes any vowel except अ, then the preceding ओ is changed to अव् and, if it is between independent words, it usually further becomes अ (i.e., the व् drops). Notice how the same rows match across, e.g., ओ row matches अव् row, and certain rows are shared, e.g., “guṇa vowel” row is shared by both ए row and ओ row.
In these listings a later rule, if equally applicable, takes precedence over a prior and thus more general rule.
In the following rules, the word “radical” means “belonging to a root or a substitute for a root”; this may apply to verbal roots made into nouns or into verbs.
May” and “usually” mean optionally infrequent or frequent change, respectively; “sometimes” and “often” mean in certain words this change is found.

2.6: General vowel sandhi.
2.7:simple vowel
(short or long)
same simple vow.
(short or long)
bothlong simple vowel
2.8:अ आsimple liquid vowel (1.9: & 2.2:)boththe liquid vowel’s guṇa (2.3:)
2.9:guṇa vowelboththe following’s component liquid vowel’s (2.3.a:) vṛddhi (2.3:)
2.10:vṛddhi vowel
2.11:simple liquid voweldifferent vowelcorresponding semivowel (2.3.a:)
2.12:guṇa vowels ए ओdrops, often replaced with an avagraha
2.13:guṇa vowelany vowel except अअय्if between indep. words (2.4:)usually→ अ a
2.15:vṛddhi vowelany vowelआय्usually→ आ a
a.The resulting vowels - अ/आ and following vowel normally remain separate, i.e., hiatus (2.1:) remains.

2.17: Special - general vowel sandhi.
2.18:final अ आ of prep. (6.11:)initial ऋ of a rootbothvṛddhi (instead of guṇa)
2.19:prep. प्रpp. ऊढbothप्रौढ “lifted up” (instead of guṇa)
2.20:past tense aug. अ (5.5.c:)initial vow. of rootbothvṛddhi (instead of guṇa)
2.21:final vowel of interj. pcl. (6.21:)
final ओ made by ॰अ + pcl. उ (6.20:)
vowelboth(no sandhi, hiatus remains)
2.22:decl. or conj. du. term. final ई ऊ ए (but not अ इ ऐ औ)
final ई of अमी (masc. pl. nom. “those” 4.9:)

2.23: Permitted finals. Except when र् precedes a क् ट् त् प्, which is at the end of a root (or a substitute for a root) and is not a suffix, no word may be permitted to end in more than one consonant. The final conjunct consonant must be reduced by dropping the last consonants until only one consonant (or क्ष्) remains. Every final consonant (whether originally a conjunct or not) must be converted into one of the eight permitted final consonants according to the following chart (2.24:). Only after this conversion can the following rules (2.24.b:, & 2.25: to 2.62:) of general visarga and general consonant sandhi be applied.

2.24: Permitted final consonants of a (conjugated, declined, or indeclinable) word.
The consonants to the left of the arrow (→) become the consonant to the right of the arrow. The column on the right indicates certain consonants which may become either of two different consonants.
क्ख्ग्घ्च्क्ज् श् ह् क्ष् → क् , or in certain words ट्
त्थ्द्ध्त्a. These apply in general sandhi combinations, so they also include finals of nominal stems before certain terminations or suffixes, as defined in 2.4:.
b. Before a pause, an original final mute (after these conversions) may optionally change to the permitted final’s respective soft nonaspirate consonant – but is rarely applied in manuscripts.
न्म्remains (न् म्)
र्स्: (visarga)
झ्ण्य्ल्व्are never finals

2.25: General visarga sandhi.
2.26:final : of 1st word in cmpd. (6.11: & 6.29:)initial क् ख् प् फ् of 2nd word in some
स् (otherwise, see 1.5.b:)
2.27:: (visarga)hardpalatal च् छ्palatal श्
cerebral ट् ठ्cerebral ष्
dental त् थ्dental स्
2.28::sibilant श् ष् स्mayश् ष् स् resp.
2.29:: after vowel
ex. अ आ
soft letter (i.e., vowel or cons.):soft र् (ex. if followed by र् per 2.35:)
2.30:आः:drops (hiatus remains)
2.31:अःvowel except अ:drops (hiatus remains)
2.32:ओ and following अ is dropped, usually replaced with avagraha ऽ (2.12:)
2.33:soft cons.
2.34:: as an etymological र्after अ आsoft lettersoft र् (exception to rules 2.30–33:)
2.35::र्drops & a preceding short vowel lengthens (exception to 2.29: & .33:)

2.36: Special - general visarga sandhi.
2.37:masc. sg. nom.
सः “he,” एषः “this” (7.12:)
any letter except अ :  → drops (hiatus remains)
2.38:भोः “O your honor” an irreg. contraction of voc. sg. of भवत् (3.13.b:)
भगोः “O virtuous one” an irreg. contraction of sg. voc. of भगवत्
soft letter
2.39:: as an etymological र्of radical noun stem (3.6.g:)pl. loc. term. सु → र् (e.g., द्वार्षु, अहर्पति)
2.40:of अहर् “day” (3.21.a:) & स्वर् “heaven”पति “lord” in cmpd.

2.41: General consonant sandhi.
2.42:soft mute cons.hard consonantits resp. hard non-aspir. cons.
hard mute cons.soft mute consonantits resp. soft non-aspir. cons.
2.43:permitted hard final क् ट् त् प् (2.24:)initial soft letterits resp. soft non-aspir. cons. (i.e., ड् द् ब् resp.), a following ह् usually → resp. soft aspir. (2.58:)
2.44:initial soft न् म्usuallyits resp. soft class nasal (else 2.43:)
2.45:hard dental त्init.hard palatal च् छ् श्hard palatal च्, a following श् usually → छ् (by 2.56:)
soft palatal ज् झ्soft palatal ज्
2.46:hard cerebral ट् ठ्hard cerebral ट्
soft cerebral ड् ढ्soft cerebral ड्
2.47:init. soft dental ल्soft द् → ल्
2.48:soft gutt. ङ्init. hard sibi. श् ष् स्mayinterpose hard gutt. क्
2.49:soft dental न्init. hard स्mayinterpose hard dental त्
2.50:init.hard palatal च् छ्anusvāra with resp. hard sibilant (श् ष् स्) interposed
hard cerebral ट् ठ्
hard dental त् थ्
2.51:init. ल्nasalized ल्, usually written as ल्ँ (1.10:)
2.52:palatal soft ज् झ् or sibilant श्palatal nasal ञ्, a following श् usually → छ् (2.56:)
2.53:init. soft cerebral ड् ढ्cerebral nasal ण्
2.54:labial म्init. semiv., sibi., or ह्anusvāra, remains म् if followed by ह्म्
2.55:mute, or nasal न् म्anusvāra → resp. class nasal opt. (necessarily in internal sandhi)
2.56:hard palatal च् (2.45:) or soft palatal ञ् (2.52:)init. श् followed by soft letterpalatal श्
hard aspir. palatal छ्
2.57:hard क् ट् प्श् may
2.58:soft ग् ड् द् ब् (2.43:)init. ह्ह् usuallyresp. soft aspir. (to घ्, etc.)
2.59:long vowel or diphthonginit. छ्opt. छ्doubles (as च्छ् c-ch). In the doubling of cons., an aspirate is doubled with its nonasp.; conj. aspir. before mute, sibi., or ह् is not allowed (2.76:, cf. 1.17:)
2.60:short vowel, or pcls. आ माछ्
2.61:vowelछ् in middle of word a
2.62:ङ् न् after short vowelinit. vowelङ् न्doubles (ङ्ङ्, न्न्)
2.63:र्single cons., ex. स् श् ष् ह्, in middle of word afollowing
in pronunciation is doubled, in writing is opt. doubled (e.g., वर्तते or वर्त्तते, दीर्घ or दीर्ग्घ)
2.64:mute preceded by nasal, or doubled mute joined with any middle of word a1st of the conj. mutesopt. drops (e.g., युङ्ग्धि or युन्धि, तत्त्व or तत्व)
a.Rule 2.61:, .63:, & .64: are actually internal sandhi, but for convenience are included in this consonant doubling/abbreviation section of rules. The doubling of a consonant in the middle of a word (2.63:) after र् is rarely followed in recent manuscripts; the abbreviation (2.64:) is not uncommon.

2.65: Internal sandhis. Being difficult to grasp without a contextual reference, they may best be learned contextually in the subsequent chapters on declensions and conjugations, in which the rules are amply exemplified and referenced. After learning those chapters, then these charts may easily be understood. Here “s.f.” stands for either nominal stem final or verbal base final, as appropriate.

2.66: Internal vowel sandhi (the most common).
2.67:s.f. इ ई, उ ऊ, ऋ
(esp. in monosyllabics
or after conj. cons.)
voweloftenइय्, उव्, इर् resp.
(instead of 2.7: & 2.11:)
2.68:s.f. ॠ term. init. cons.ईर्
2.69:s.f. ॠ after labialऊर्
2.70:s.f. ऋ after sing. con.term. init. य् (5.37.3:)रि
2.71:s.f. ए ऐ औterm. init. vowel or य्अय् आय् अव् आव् resp.
(compare & constrast 2.13: to 2.16:)

2.72: Internal consonant sandhi (the most common).
2.73:s.f. cons.term. init. vow., semiv., nasalusuallyfollows internal, but not general sandhi
2.74:term. init. ह्, sibi., muteusuallyfollows internal, then general sandhi
2.75:final sing. cons. term.term.drops (cf. 2.23:), final s.f. cons. then follows general sandhi
2.76:s.f. aspirateह्, sibi., mute, or via 2.24:aspiration is lost
2.77:(lost) soft aspiration (2.76:)ध्व् भ् स्if
aspiration thrown backward (cf. 2.79:)
2.78:term. init. त् थ्aspiration thrown forward on softened त् थ् (i.e., → ध्)
2.79:initial ग् द् ब् of radical syllableradical final is aspirate घ् ध् भ् ह्if final aspir.
is lost
initial ग् द् ब् is aspirated (cf. 2.77:)
2.80:s.f. palatal च्hard or soft cons.usuallygutt. hard क् or soft ग् resp. (cf. 2.24:)
2.81:s.f. palatal ज्hard क् or soft ग्, sometimes hard ट्/ष् or soft ड् resp. (cf. 2.24:)
2.82:s.f. palatal च् ज्vowel, semiv., nasalsome-
gutt. ग् (cf. 2.24: & 2.42:), sometimes क्
2.83:s.f. palatal छ् of प्रछ् “ask”cons., ex. pada term.श् (then 2.84–86: may apply)
2.84:s.f. palatal श्hard nom. स्, or a soft ध् or pada term.hard ट् or soft ड्, sometimes hard क् or soft ग् resp. (cf. 2.24:)
2.85:त् थ्ष्, and then both → ष्ट् or ष्ठ् resp. (2.88:)
2.86:स् of conj. term.क् (cf. 2.24:), and then both क्ष् (2.101:)
2.87:s.f. palatal च् ज्न्न्palatal ञ् (cf. 2.73:)
2.88:s.f. cerebralhard or soft dental class cons.dentalhard or soft cerebral resp. (cf. 2.73:)
2.89:s.f. cerebral ष्hard or soft decl. term. init. cons.,
or ध् conj. term. init.
hard ट् or soft ड् resp. (cf. 2.24: & .42:), a following dental → cerebral (2.88:)
2.90:conj. term. init. स्क्, and then both → क्ष् (2.101:)
2.91:root final क्ष्term. init. mute or sibi.क्drops, leaving ष् to combine as above
2.92:cerebral ऋ ॠ र् ष् (w/wo an intervening a vowel, guttural incl. ह्, labial incl. व्, य्, or anusvāra)न्, if followed by a vowel or न् म् य् व्न्ण् (a following conj. न् म् also → ण्).
(This rule also applies between many cmpd. words 6.29: (ex. only sometimes when न् is in conjunct with preceeding gutt. or labial cons.), including between all prefixes and verbal forms or nouns 6.11:)
2.93:root final न्स्anusvāra (cf. 2.54:)
2.94:न् at end of, or inserted in (3.2.b:), stemष् स् ह्
2.95:न्muteresp. class nasal
2.96:स् of वस् “dwell”
& घस् “eat”
init. स् of conj. term.त्, softening it to द् before soft भ्
2.97:s.f. स् of redup. perf. (3.19:)term. init. भ् or स्, and neut. sg. nom. acc.
2.98:स्between mutesdropsl
2.99:prep. उद् (in ext. sandhi)roots स्था “stand”
& स्तम्भ् “support”
स्drops (2.98:), उद् changes to hard उत् (2.42:)
2.100:स्soft dentalष् → drops, following dental may → cerebral (2.88:)
2.101:vowels ex. अ आ b (w/wo intervening anusvāra or : visarga ),
or क् र्
स् if followed by vow.
or त् थ् न् म् य् व्.
(But a following root or derivative with र् or ऋ, whatever its position, usually blocks the change.)
स्ष्, the following dental त् थ् न् becoming cerebral also.
(This rule also applies between many compound words 6.29:, including between all prefixes and verbal forms or nouns 6.11:)
2.102:म्term. init. व्न्
2.103:semi-vowels य् र् ल्remains म् (not → anusvāra per 2.54:)
2.104:sibilant or ह् in the middle of a wordanusvāra (2.54:)
2.105:soft ह्स्घ् (cf. 2.24:), then both → क्ष् (2.76:, .42:, .101:)
2.106:final ह् of roots w/initial द्term. init. त् थ् ध्घ् (then 2.76:–2.78:)
2.107:final ह् of roots wo/init. द्term. init. dental
त् थ् ध्
ढ् (cf. 2.24:) → drops, & dental → asp. ढ् (2.78: & .88:), w/preceding vow. → long (e.g., लिह्+त→लीढ)
2.108:स्निह् “be oily” & sometimes मुह् “be confused”pp. term. तह्घ्, then both → ग्ध् (2.76:, .78:)
2.109:ह् of वह् “convey” & सह् “endure”term. init. dental
त् थ् ध्
ढ् → drops, and dental → asp. ढ् (2.78: & 2.88:),
with preceding अ → ओ
2.110:root वह् “convey”pp. term. तbothऊढ (by saṃprasāraṇa 2.3.a: & .107:)
2.111:root दृह् “be strong”दृढ (2.107:)
2.112:root नह् “bind”नद्ध
a.Cerebralization (2.92:) of a distantly following न् (nasals being very susceptible to change of location/pronunciation-position) is allowed to occur when no intervening sound either “satisfies the need” to follow up the preceding non-touched (1.4:) cerebral (i.e., by an intervening cerebral mute) or moves the front part of the tongue out of position to cerebralize the nasal (i.e., by an intervening dental, sibilant or palatal - except the weak, i.e., slightly touched [1.4:] palatal य्).
b.The sound ष् hardly occurs of its own in words, apart from application of this rule (2.101:) and, to a lesser extent, rules 2.81:, .85: and .86:. In other words, if you see a ष्, suspect it was a स्.

Go to Top ('Home' key on many Windows browsers)

3.1–3 3.4 3.5–6
3.7–9 3.10–1 3.12–3 3.14 3.15 3.16
3.17–8 3.19 3.20–4 3.24–5 3.26
3.27 3.28 3.29–30 3.31−5 3.36 3.38–40 3.41

The Inflection of Nominal Stems by Means of Case Terminations

3.1: Inflection. The inflection of nouns (including adjectives), numerals and pronouns is called declension (विभक्ति). In Saṃskṛta declensions there are-
a. Three genders (लिङ्ग-s): masculine, feminine and neuter (पुम् स्त्री नपुंसक respectively, abbreviated as पुं॰ स्त्री॰ नपुं॰). See 6.26:−.28: for examples of noun stems and gender.
b. Three numbers (वचन-s): singular, dual and plural (एक द्वि बहु respectively, abbreviated as ए॰ द्वि॰ ब॰). See 7.4:−.5: for discussion of number.
c. Eight cases (विभक्ति-s): nominative, vocative, accusative, instrumental, dative, ablative, genitive and locative. This is also the native grammarian order except that vocative, which, not being considered a proper case by them, is ordered at the end. I have followed the non-native order, since this shows how closely the vocative follows the nominative, making it easier to remember. The native grammarians name these seven cases - प्रथमा द्वितीया तृतीया चतुर्थी पञ्चमी षष्ठी सप्तमी i.e., 1st – 7th, see 4.26:, abbreviated as प्र॰ द्वि॰ तृ॰ च॰ प॰ ष॰ स॰. They call the vocative सम्बोधन (सं॰). See 7.7:−.13: for discussion of the cases.

3.2: Normal Case Terminations. This table should be committed to memory as it forms the basis for nominal declensions of most words, especially the consonant-ending declensions.
Generally only in the first three cases do the neut. terms. differ from the masc. and fem., hence they are shown separately below only for those first three cases.
These terminations are added to a nominal stem, called a प्रतिपादिक, to form a declined word fit to be employed in a sentence. Hyphens (—) indicate an absence of a termination, and hence, in those declined cases, there would appear no distinction from their प्रतिपादिक.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m. स्त्री॰ f.नपुं॰ n.पुं॰ m. स्त्री॰ f.नपुं॰ n.पुं॰ m. स्त्री॰ f.नपुं॰ n.
प्र॰ N.स्अस् b
सं॰ V.
द्वि॰ A.अम्
तृ॰ I.भ्याम् aभिस्
च॰ D.भ्यस्
प॰ Ab.अस्
ष॰ G.ओस्आम्
स॰ L.सु
a.भ्याम् भिस् भ्यस् सु are called pada terminations.
b.N. pl. nom. voc. acc. insert न् before a single final mute (changing to resp. class nasal 2.95:), sibilant or ह् (2.94:) of a consonant-ending stem, and insert न् after a vowel-ending stem.

3.3: Accent. In declension, accents follow the below general rules.
1. Vocative case is always accented on the first syllable.
2. Monosyllabic stems have their accent move to the weak (3.17:) or middle (pada) terminations.
  • Root-words in long ई ऊ (ī̍ ū̍) as final of cmpd. (6.24: & .38:) retain their accent throughout.
3. Accent of some polysyllabic stems ending in consonant moves to the weak termination.
  • Present participle stems ending in accented अत् a̍t (3.12:).
  • When the accented vowel in the final suffix is dropped, such as by syncopation (3.20.b:).
  • A few other sporadic instances.
4. Polysyllabic stems ending in an accented vowel shift the accent to a vowel-initial termination, if the accented stem vowel is lost by replacement with a semi-vowel (2.2:, 2.67:–.71:).
  • The genitive pl. term. आम् (3.29:–.30:, 3.38:) may also take the accent even though the nasal न् is inserted between it and the short accented इ उ ऋ (i̍ u̍ or ṛ̍).

3.4: Consonant-ending Stems. Nouns end in either consonants or vowels. The consonant-ending declensions are more regular than the vowel-ending, so we will present them first. Consonant-ending nouns can be divided into unchangeable stems and changeable stems.
Although native grammarians for memorization teach recitation across the row (nom. sg. du. pl…), the visual and logical patterns go down the column (sg. nom. voc. acc…).

3.5: Unchangeable Stem – Dental ॰त्मरुत् (maru̍-t) m. “wind”
जगत् (ja̍ga-t) n. “world”
m. मरु॰
n. जग॰
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m. स्त्री॰ f.नपुं॰ n.पुं॰ m. स्त्री॰ f.नपुं॰ n.पुं॰ m. स्त्री॰ f.नपुं॰ n.
प्र॰ N.॰त् (स्) a॰त् (—)॰तौ (औ)॰ती (ई)॰तः (अस्)॰न्ति (इ) d
सं॰ V.॰त् (—)
द्वि॰ A.॰तम् (अम्)
तृ॰ I.॰ता (आ)॰द्भ्याम् (भ्याम्) c॰द्भिः (भिस्)
च॰ D.॰ते (ए)॰द्भ्यः (भ्यस्)
प॰ Ab.॰तः (अस्) b
ष॰ G.॰तोः (ओस्)॰ताम् (आम्)
स॰ L.॰ति (इ)॰त्सु (सु)
a.Rule 2.23:, final cons. (स्) after another cons. drops, hence मरुत् + स् = मरुत्स् becomes मरुत्.
b.Rule 2.24:, the final स् becomes : (visarga).
c.भ्याम् being a pada term. (3.2.a:), general (2.4:) sandhi rule 2.43: applies, hence त् is softened.
d.Rule 3.2.b:, न् is inserted b/4 a single final mute (त्) in n. pl. nom. voc. acc., ∴ jaga-n-t-i.

3.6: The rest of the unchangeable stems. It is only necessary to show four forms that exemplify the various formations during declension. The rest of the forms can be easily extrapolated from these four based on the nature of what follows the stem, then adding the appropriate normal case terms. Exceptional forms are dark-shaded.
sg. nom.
॰Vow. term.
pl. nom.
m./f.॰अः / n.॰इ
॰Soft Pada
pl. inst.
॰Hard Pada
pl. loc.
॰म-थ् -thin cmpd. a.
॰त् -t a॰थः m.f.
॰न्थि n. -nth
॰द्भिः -d॰त्सु -t
॰सुहृ-द् -dm. “friend”॰त् -t॰दः॰द्भिः -d॰त्सु -t
॰वृ-ध् -dhin cmpd. a.
॰त् -t॰धः m.f.
॰न्धि n. -ndh
॰द्भिः -d॰त्सु -t
Palatalवा-च् -cf. “speech”॰क् -k b॰चः॰ग्भिः -g॰क्षु -kṣ c
असृ-ज् -jn. “blood”॰क् -k॰ञ्जि n. -ñj d॰ग्भिः -g॰क्षु -kṣ
रु-ज् -jf. “disease”॰क् -k॰जः॰ग्भिः -g॰क्षु -kṣ
सम्रा-ज् -jm. “sovereign”॰ट् -ṭ e॰जः॰ड्भिः -ḍ॰ट्सु -ṭ
दि-श् f. “cardinal point”॰क् -k॰शः॰ग्भिः -g॰क्षु -kṣ
वि-श् m. “settler”॰ट् -ṭ॰शः॰ड्भिः -ḍ॰ट्सु -ṭ
Cerebralद्वि-ष् -ṣm. “enemy”॰ट् -ṭ॰षः॰ड्भिः -ḍ॰ट्सु -ṭ
Labialधर्मगु-प् -pm. “guardian
of dharma
॰प् -p॰पः॰ब्भिः -b॰प्सु -p
ककु-भ् -bhf. “region”॰प् -p॰भः॰ब्भिः -b॰प्सु -p
(only र् 2.24)
द्व-ार् -ārf. “gate”॰आः -āḥ॰आरः॰आर्भिः -ār f॰आर्षु -ārṣ g
गिर् -irf. “voice”॰ईः -īḥ h॰इरः॰ईर्भिः -īr॰ईर्षु -īrṣ
पुर् -urf. “town”॰ऊः -ūḥ॰उरः॰ऊर्भिः -ūr॰ऊर्षु -ūrṣ
Stems in स्यशस् -asn. “fame”॰अः -aḥ॰आंसि n. -āṃs i॰ओभिः -o j॰अःसु -aḥ k
हविस् -isn. “oblation”॰इः -iḥ॰ईंषि n. -īṃṣ॰इर्भिः -ir॰इःषु -iḥṣ
आयुस् -usn. “life”॰उः -uḥ॰ऊंषि n. -ūṃṣ॰उर्भिः -ur॰उःषु -uḥṣ
सुमनस् -asa. “cheerful”॰आः m.f.-āḥ i
॰अः n. -aḥ
॰असः m.f.
॰आंसि n. -āṃs
॰ओभिः -o॰अःसु -aḥ
आशिस् -isf. “blessing”॰ईः -īḥ l॰इषः -iṣ॰ईर्भिः -īr॰ईःषु -īḥṣ
Stems in ह्॰दुह् -duhin cmpd. a.
॰धुक् -dhuk m ॰दुहः m.f.
॰दिंहि n.
॰धुग्भिः -g m॰धुक्षु -kṣ m
॰लिह् -lihin cmpd. a.
॰लिट् -ṭ॰लिहः m.f.
॰लिंहि n. -ṃh
॰लिड्भिः -ḍ॰लिट्सु -ṭ
a.Rule 2.23:, final conjunct reduces to initial cons., hence ॰मथ् + स् = ॰मथ्स् becomes ॰मथ्, then rule 2.24:, final थ् becomes त्.
b.Rule 2.24:, final च् and sometimes श् becomes क्.
c.Rule 2.24.a:, final च् becomes क्, then rule 2.101:, because of preceding क्, स् of सु becomes ष्, and they both are written क्ष्.
d.Rule 2.95:, inserted न् (3.2.b:) before the soft palatal mute (ज्) in n. pl. nom. becomes palatal ञ्.
e.Rule 2.24:, final ज् श् ह् in certain words becomes ट्.
f.Rule 2.34:, visarga (:) representing etymological र् after अ or आ reverts back to र्.
g.Rule 2.39:, final र् is retained before pl. loc. term. सु, the latter then becoming षु (2.101:).
h.Stems in इर् and उर् lengthen the preceding इ or उ in sg. nom. and before pada terms.
i.Stems in स् are almost all from the primary nominal suffixes अस् इस् उस्, chiefly neuters. They lengthen their final vowel before the inserted nasal (changed to anusvāra 2.94:) in n. pl. nom. voc. acc., but not before other vowel initial terms. e.g., n. sg. inst. यशसा, हविषा, आयुषा (2.101:), etc. The masculine and feminine are nearly all adjectival compounds as well as proper names like अङ्गिरस्; those in अस् lengthen the अ in m.f. sg. nom.
j.Rule 2.33:, अ plus visarga (:) change to ओ before a soft consonant.
k.Rule 2.28:, यशःसु may also be written यशस्सु.
l.आशिस् lengthens its इ (like stems in इर् above 3.6.h:) in sg. nom. and before pada terminations.
m.Rule 2.24:, final ह् usually becomes क्, and this lost aspiration is thrown on root initial ग् द् ब् (2.79:).

Consonant-ending Changeable Stems

3.7: Changeable stems. They also take the normal case-endings. The stems exhibit a strong, a weak and sometimes a third form, called middle. The stems having been usually accented in the strong cases preserved their full form, but were shortened in the weak cases by the accent often falling on the terminations. Similarly, if the strong stem has a long vowel within it, that long vowel is regularly shortened (i.e., weakened) in sg. vocative because of the accent shifting in that case to the first syllable.
The consonant-ending changeable stems are generally listed with their middle (or weak) form (the form before the pada terminations), which is the form normally employed in compounds (6.29:) and listed in dictionaries. Whereas, the unchangeable consonant-ending stems as well as the vowel-ending stems, introduced later in this chapter, show their प्रतिपादिक form (3.2:) in dictionary and in compounds.

3.8: General Rules for Changeable Stems.

1. The vowel of the suffix is lengthened in the strong m. sg. nom., except in the stems अत् and अच्.
2. The changeable stems that originally end in न् drop it in m. sg. nom. The other changeable stems insert a न् in their strong stem, but in m. sg. nom. via rule 2.23: that inserted न् remains as the final.
3. All changeable stems which lengthen their vowel in the strong m. sg. nom. or because of 3.8.1:, in vocative keep it short and always retain the न् of that strong stem.
  • In other changeable stems, the vocative differs from the nominative only in having the accent always on the first syllable.

3.9: Feminines of changeable stems are formed by adding ई to the weakest form of the stem retaining the final consonant, and are declined like the polysyllabic नदी (3.36:).
  • For example, ददती dad-at-ī, श्रेयसी śre-yas-ī, धनिनी dhan-in-ī (strong), विदुषी vid-uṣ-ī, राज्ञी rāj-ñ-ī, नाम्नि nām-n-ī, प्रतीची prat-īc-ī.
Also, the feminines of the uncommon unchangeable consonant adjectival stems, except of bare roots (6.24:) and stems in स् (3.6.i:) (e.g., the adjectives ॰मथ् and सुमनस् in 3.6:), are formed by adding ई, and are declined like the polysyllabic नदी.
a. The feminine of pr. pt. active, and usually the simple fut. pt. active, (stems in अत् see 6.4:) differ when formed from first conj. roots (5.2:), where they add the ई to the strong form of the stem, i.e., अन्ती.
b. The few adjectives in ॰वन् (3.22:) form their feminine in ॰वरी. The feminine of the irreg. युवन् “young” (3.23:) is युवति or युवती.

3.10: Two-Form Stems with the Normal Case Terminations. The strong cases are dark-shaded; the rest are weak, and have the same terminations as in 3.2:.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.स्अस्
सं॰ V.
द्वि॰ A.अम्अस्
. . .
a.Forming an exception to this are the stems in ॰इन्, ॰मिन्, and ॰विन् (3.15:). There, all the cases are strong, except those of the weak pada terms. and the n. sg. nom. acc. and optionally the voc.

3.11: Two-Form Stems.
Part of speechPara.#SuffixStrong FormWeak Form
Pr. & Fut. Pt. act. (6.4:)3.12:॰अत्॰अन्त्॰अत्
Poss. Adj. & Subst. (6.27:)3.13:॰मत्॰मन्त्॰मत्
Poss. Adj. & Subst. (6.27:)
& Past Active Pt. (6.7:)
Comparative (6.25:)3.14:॰(ई)यस्॰(ई)यांस्॰(ई)यस्
Poss. Adj. & Subst. (6.27: & .26:)3.15:॰इन्॰इन्॰इ

3.12: Two-Form Stems – ॰अत्(strong ॰अन्त् a, weak ॰अत्)
(pr. pt. act. w/strong in ॰अत् 6.4: if redup. decline like ॰त् 3.5:, opt. inserting न् in n. pl nom. voc. acc. 3.2.b:)
गछत् (gacch-at) a. “going”
Pr. & Fut. Pt. act.
a. गच्छ्॰ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.॰अन् (स्) b॰अत् (—)॰अन्तौ (औ)॰अती (ई)॰अन्तः (अस्)॰अन्ति (इ) d
सं॰ V.॰अन् (—) c
द्वि॰ A.॰अन्तम् (अम्)॰अतः (अस्) d
तृ॰ I.. . .. . .॰अद्भिः (भिस्)
. . .
स॰ L.. . .. . .॰अत्सु (सु)
a.महत् “great,” originally a pr. pt., declines like 3.12: ex. forms strong stem in ॰आन्त्, sg. voc. is ॰अन् (3.8.3:).
b.Rule 3.8.1:, the stem in ॰अत् is an exception to this rule of lengthening the vowel in the strong m. sg. nom.
c.Rule 2.23:, final त् of strong stem ॰अन्त् is dropped.
d.The rest of the weak forms can be easily determined from the example in weak m. pl. acc.

3.13: Two-Form Stems – ॰मत्/॰वत्(s. ॰मन्त्/॰वन्त्, w. ॰मत्/॰वत्)
धीमत् (dhī-m-at) a. “intelligent”
भगवत् (bhaga-v-at) m. “the Lord”
Possessive Adj. & Subst. (॰मत् / ॰वत्)
& Past Active Pt. (॰वत्)
a. धी॰
m. भग॰
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.॰मान्/वान् (स्) a. . . the rest like (म्॰/व्॰) अत् (3.12) . . .
a.Rule 3.8.1:, the vowel is lengthened in the strong m. sg. nom.
b.भवत् (bha̍v-at), as pr. pt. of √भू (i.e., the root भू), meaning “being” is declined like ॰अत् (3.12:). As a pronoun meaning “your honor,” it declines like ॰वत् (cf. 4.18:), i.e., m. sg. nom. भवान्, du. भवन्तौ, etc.
c.कियत् (ki-yat) “how much?” and इयत् (i-yat) “so much” (4.18:) decline like ॰वत्, e.g., m. sg. nom. कियान्, du. कियन्तौ, etc.

3.14: Two-Form Stems – ॰(ई)यस् a(s. ॰(ई)यांस्, w. ॰(ई)यस्)
श्रेयस् (śre̍-yas) a. comparative “better”
a. श्रे॰ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.॰(ई)यान् (स्)॰(ई)यः (—)॰(ई)यांसौ (औ)॰(ई)यसी (ई)॰(ई)यांसः (अस्)॰(ई)यांसि (इ)
सं॰ V.॰(ई)यन् (—)
द्वि॰ A.॰(ई)यांसम् (अम्)॰(ई)यसः (अस्)
तृ॰ I.. . .. . .॰(ई)योभिः (भिस्)
. . .
स॰ L.. . .. . .॰(ई)यःसु (सु)
a.ई of ॰ईयस् (see 6.25:) is a connecting vowel that is sometimes dropped when forming the stem, e.g., श्रेयस् śre̍-yas.

3.15: Two-Form Stems – ॰इन्/॰मिन्/॰विन्(s. ॰इन्/॰मिन्/विन्, w. ॰इ/॰मि/॰वि)
धनिन् (dhan-in) a. “wealthy”
स्वामिन् (svā-m-in) m. “master”
Possessive. Adj. & Subst.
a. धन्॰
m. स्वा-म्॰
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.॰ई (स्) a॰इ (—)॰इनौ (औ)॰इनी (ई)॰इनः (अस्)॰ईनि (इ) c
सं॰ V.॰इन् (—)॰इन्/॰इ (—) b
द्वि॰ A.॰इनम् (अम्)॰इ (—)
तृ॰ I.॰इना (आ)॰इभ्याम् (भ्याम्)॰इभिः (भिस्)
च॰ D.॰इने (ए)॰इभ्यः (भ्यस्)
प॰ Ab.॰इनः (अस्)
ष॰ G.॰इनोः (ओस्)॰इनाम् (आम्)
स॰ L.॰इनि (इ)॰इषु (सु)
a.Rule 3.8.1:, stem vowel is lengthened in m. sg. nom. Rule 3.8.2:, in m. sg. nom. न् is dropped for stems ending in न्.
b.Refer to 3.10.a: for this irreg. distribution of strong and weak forms. The n. sg. voc. has an alternate weak form, like the n. sg. nom. acc.
c.These stems in n. pl. nom. voc. acc., like m. sg. nom., lengthen the stem vowel.

3.16: The irregular word अप् f. “water(s)” has a strong and a weak form, and only declines in plural.
ए॰ Singularद्वि॰ Dualब॰ Plural
प्र॰ N.आपः (अस्) a
सं॰ V.
द्वि॰ A.अपः (अस्)
तृ॰ I.अद्भिः (भिस्) b
. . .. . .
स॰ L.॰अप्सु (सु)
a.The strong pl. nom. voc. lengthen the अ.
b.Substitute त् for final प् before terminations in भ्, which then softens (2.42:)

3.17: Three-Form Stems with the Normal Case Terminations. The strong cases are dark-shaded; the middle (before pada terms. and n. sg. nom. voc. acc.) are light-shaded; the rest are weak.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.स्अस्
सं॰ V.
द्वि॰ A.अम्अस्
तृ॰ I.भ्याम्भिस्
च॰ D.भ्यस्
प॰ Ab.अस्
ष॰ G.ओस्आम्
स॰ L.सु

3.18: Three-Form Stems.
Part of speechPara.#SuffixStrong
Redup. Past Active Pt. (6.4:)3.19:॰वस्॰वांस्॰वत्॰उष्
Mostly masc. agent nouns (6.26:)3.20:॰अन्॰आन्॰अ॰न्
Mostly neuter action nouns (6.26:)3.22:॰मन्॰मान्॰व॰म्न्
Possessive adj. (6.27:)
and masc. & neuter subst. (6.26:)
Adj. expressing “-ward,” formed
by cmpd. with root अञ्च् “bend”
3.24:॰अच्॰अञ्च्॰अच्॰ईच् after य् a
॰ऊच् after व्
a.Some words with the stem ॰अच् only have two forms – a strong and weak, the weak taking the form of ॰अच्. Like with the two-form stems, in these words the weak form is taken by all cases that are not strong (see 3.25:).

3.19: Three-Form Stems – ॰वस् a(s. ॰वांस्, m. वत्, w. ॰उष् a)
विद्वस् (vid-va̍s) (5.28: & 6.4:) a. “wise”
Reduplicative Perfect Participle
a. विद्॰ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.॰वान् (स्)॰वत् (—)॰वांसौ (औ)॰उषी (ई)॰वांसः (अस्)॰वांसि (इ)
सं॰ V.॰वन् (—)
द्वि॰ A.॰वांसम् (अम्)॰उषः (अस्) b
तृ॰ I.. . .. . .॰वद्भिः (भिस्)
. . .
स॰ L.. . .. . .॰वत्सु (सु)
a.The व is weakened to उ, called saṃprasāraṇa (2.3.a:). Then rule 2.101:, under influence of preceding liquid vowel (or क् र्), स् becomes ष्.
b.Certain of these participles insert a connecting vowel इ before the suffix ॰वस्. In the weak form, where the stem turns to ॰उष्, instead of a further confusion of the original by application of vowel sandhi between इ and उ (→यु), the preceding इ is dropped, e.g., तस्थिवस् (tasth-i-va̍s) “that which has stood” in the weak cases would be तस्थुष्॰ (tasth-uṣ) plus the weak vowel terminations.

3.20: Three-Form Stems – ॰अन्(s. ॰आन्, m. अ, w. ॰न्, opt. weak ॰अन्)
राजन् (rā̍j-an) m. “king”
Mostly masc. agent nouns
m. राज्॰ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.पुं॰ m.पुं॰ m.
प्र॰ N.॰आ (स्) a॰आनौ (औ)॰आनः (अस्)
सं॰ V.॰अन् (—)
द्वि॰ A.॰आनम् (अम्)राज्ञः (अस्) b
तृ॰ I.. . .. . .॰अभिः (भिस्)
. . .
स॰ L.राज्ञि b or राजनि c (इ). . .॰असु (सु)
a.Rule 3.8.1:, stem vowel is lengthened in m. sg. nom. Rule 3.8.2:, in m. sg. nom. न् is dropped for stems ending in न्.
b.Rule 2.87:, in the weak forms of राजन्, the preceding palatal ज् converts the weak न् to palatal ञ्, both written with the conjunct ज्ञ्. If the original अन् was accented (a̍n), then the lost accent in weak form is thrown forward onto the termination (3.3.3:)
c.In weak sg. loc, instead of ॰न्, the optional stem ॰अन् may be applied.

3.21 Irregular forms of ॰अन्.
अहन्(a̍h-an) n. “day” (6.26:)॰आन्॰अस् aन्
पन्थन्(pa̍nth-an) m. “path”पन्थान् bपथिपथ्
श्वन्(ś-v-a̍n) m. “dog”॰वान्॰वउन्
॰हन् c(-h-an) a. at end of cmpd. “destroyer”॰हन्॰हघ्न्
a.अहन् has the normal distribution of three-form stems in neuter (3.17:), hence the middle forms would be sg. nom. voc. acc. अहः, pl. inst. अहोभिः (2.24: & 2.33:) etc., and pl. loc. अहःसु (2.24:). This middle stem अहस् indicates that अहन् is actually a defective stem, needing to borrow that form from another word, अहर्/अहस् which also means “day,” to complete its full declension. As a prior member of a compound, where the middle stem is normally employed (6.29:), this ॰अस् is almost always treated as ॰अर्, e.g., अहरहः (ahar-ahaḥ 2.34:) “day by day.”
b.Rule 3.8.2: applies to the m. sg. nom. of this and the following two irregular stems where final न् is dropped. For पन्थन्, after dropping final न्, the sg. nom. and the sg. voc. then add स् to form पन्थाः. Only the strong form retains the medial nasal.
c.हन् is a root used at the end of a cmpd. (6.24:), but is declined similar to a stem in ॰अनु. It also takes long vowel in sg. nom. only (॰हा) and allows the optional stem हन् h-an in weak sg. loc.

3.22: Three-Form Stems – ॰मन्/॰वन्(s. ॰मान्/॰वान्, m. ॰म/॰व, w. ॰म्न्/॰व्न्,
opt. w. ॰मन्/॰वन्)
आत्मन् (āt-ma̍n) m. “self”
नामन् (nā̍-man) n. “name”
Mostly neuter action nouns (॰मन्)
& Possessive a., m. & n. subst. (॰वन्)
n. ना॰
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.॰मा (स्) a॰म (—)॰मानौ (औ)॰म्नी or ॰मनी b (ई)॰मानः (अस्)॰मानि (इ)
सं॰ V.॰मन् (—)॰म or ॰मन् b (—)
द्वि॰ A.॰मानम् (अम्)॰म (—)॰म्नः or ॰मनः c (अस्)
तृ॰ I.॰म्ना or ॰मना c (आ). . .॰मद्भिः (भिस्)
. . .
स॰ L.॰म्नि or ॰मनि b (इ). . .॰मसु (सु)
a.Rule 3.8.1:, stem vowel is lengthened in m. sg. nom. Rule 3.8.2:, in m. sg. nom. न् is dropped for stems ending in न्.
b.Similar to 3.20.c:, In weak m.n. sg. loc, in weak n. du. nom. voc. acc., and in middle n. sg. voc. the optional stem ॰मन्/॰वन् may be applied. The rest of the weak forms do not take the optional stem ॰मन्/॰वन् unless required by 3.22.c: below.
c.The weak forms of ॰मन्/॰वन्, when preceded by a consonant, as in the example here (आत्-मन्), do not drop the stem vowel, thus here avoiding the conjunction of three consonants, here of त् t with म्न् mn. So they necessarily take the optional stem ॰मन्/॰वन्}.

3.23 Irregular forms of ॰वन्.
Strong StemMiddle StemWeak Stem
मघवन्(magh-a̍-van a) m. “a name of Indra॰अवान्॰अव॰ओन् b
युवन्(y-u̍-van a) m. “youth”॰उवान् b॰उव॰ऊन् c
a.The vowel before the ॰वन् becomes involved in the declension, so I have included it in the paradigm. Rule 3.8.2:, in sg. nom. न् is dropped for these irregular stems ending in न्.
b.The weak stem ॰ओन् is formed by saṃprasāraṇa (2.3.a:) where व weakens to उ, which then combines with the preceding व into ओ (2.8:).
c.The weak stem ॰ऊन् is formed similar to 3.23.b:, where व weakens to उ, which then combines with the preceding उ into ऊ (2.7:).

3.24: Three-Form Stems – ॰अच्(s. ॰अञ्च्, m. ॰अच्, w. ॰ईच् or ॰ऊच्) a
प्रत्यच् (prat-i-a̍c) a. “backward, westward”
Adj. “-ward,” formed by cmpd. with
verb अञ्च् “bend”
a. प्रत्ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.पुं॰ m.नपुं॰ n.
प्र॰ N.॰यङ् (स्) b॰यक् (—) c॰यञ्चौ (औ)॰ईची d (ई)॰यञ्चः (अस्)॰यञ्चि (इ) f
सं॰ V.॰यङ् (—)
द्वि॰ A.॰यञ्चम् (अम्)॰ईचः (अस्)
तृ॰ I.. . .. . .॰यग्भिः (भिस्) e/a>
. . .
स॰ L.. . .. . .॰यक्षु (सु) e
a.These adjectives normally have a य् (=इ) or a व् (=उ) before ॰अच्. In this example the first member is the preposition प्रति “back.” This इ or उ becomes involved in the declension, so it has been included in the paradigm. For those with उ b/4 ॰अच्, m. sg. nom. would be ॰वङ्, etc.
b.प्रत्यञ्च्स् is reduced to प्रत्यञ् by 2.23: and then is changed to permitted final प्रत्यङ् by 2.24:.
c.The middle stem प्रत्यच् is changed to permitted final प्रत्यक् by 2.24:.
d.The य् or व् before ॰अच् weakens to ॰ईच् or ॰ऊच् respectively in the weak cases.
e.The middle stem final च् is changed to permitted final क् before pada term., then is softened to ग् by 2.43: before soft cons. or combined with hard cons. सु term. to क्षु by 2.101:.
f.Rule 2.95:, inserted nasal न् changes to corresponding nasal of the class consonant च्.

3.25 Similarly declined ॰अच् stems.
Strong StemMiddle StemWeak Stem
अन्वच्(an-v-a̍c) “following”॰वञ्च्॰वच्॰ऊच्
न्यच्(n-y-a̍c) “downward”॰यञ्च्॰यच्॰ईच्
विष्यच्(viṣ-v-ac) “all-pervading”॰वञ्च्॰वच्॰ऊच्
सम्यच्(sam-y-a̍c) “going together”॰यञ्च्॰यच्॰ईच्
उदच्(u̍d-a̍c) “upward, northern”॰अञ्च्॰अच्॰ईच् a
तिर्यच्(tir-y-a̍c) “traverse”॰यञ्च्॰यच्॰अश्च् b
Exceptions with only two-forms (3.18.a:)
अवाच्(a̍v-a-ac) “downward, southern”॰आञ्च्॰आच्
पराच्(pa̍r-a-ac) “turned away”
प्राच्(pr-a̍-ac) “forward, eastern”
a.This word has no य् but still takes ॰ईच् as if it did.
b.The unusual weak form is actually closer to the original first member of the cmpd., तिरस् “trans-,” the स् becoming palatal श् due to palatal च् (2.24: and .27:).

3.26 Irregular noun with a three-form stem.
Strong StemMiddle StemWeak Stem
पुमंस् or पुम्(pu̍-maṃs) m. “man”॰पुमांस् a॰पुम् b॰पुंस्
a.Sg. nom. पुमान् (anusvāra reverting to न्), sg. voc. पुमन् (3.8.3:).
b.Pl. inst. पुम्भिः, pl. loc. पुंसु (2.54:).

Vowel-ending Noun Stems

3.27: Vowel-ending noun stems. Nouns (including adjectives) ending in vowels are the most common words in Saṃskṛta. They do not take the full set of normal case terminations (3.2:), but rather a modified subset of these terminations. Even within the various vowel-ending stems, there is as much deviation as there is affinity between them. The deviation is so much that it is hardly worth trying to remember each of the termination sets in isolation, apart from the vowel stem to which they attach.
Therefore I have chosen to present the full (sandhi applied) final form of the declensions with both the final stem vowel and the terminations as a unit. In this way it is a bit easier to remember and much easier to apply to their intended stems.
Another device I have used to simplify the charts is to show just those forms in neuter and feminine that differ (noted with the mathematical symbol , called “delta” – meaning “difference”) from the masculine (or polysyllabic from mono- in 3.36:), and therefore need to be remembered. For example, ॰अ ending n. voc. sg. would be ॰अ, the same as m. voc. sg.; ॰आ ending f. nom. pl. would be ॰आः, the same as m. nom. pl.

3.28: Vowel Stems – ॰अ/॰आ aराम (rām-a) m. “Lord Rāma
फल (pha̍l-a) n. “fruit”
सीता (si̍ta-ā) f. “Sītā, wife of Rāma”
(exceptions to the normal case terms. 3.2: or the common pronominal terms. 4.4: are in bold)
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
प्र॰ N.॰अः॰अम्॰आ॰औ॰ए c॰ए c॰आः॰आनि c & e. . . c
सं॰ V.॰अ. . .॰ए b
द्वि॰ A.॰अम्. . .॰आम् c॰आन् c॰आः c
तृ॰ I.॰एन c. . .॰अया c॰आभ्याम् c. . .. . .॰ऐः c. . .॰आभिः c
च॰ D.॰आय. . .॰आयै d॰एभ्यः c. . .॰आभ्यः c
प॰ Ab.॰आत्. . .॰आयाः d
ष॰ G.॰अस्य c. . .॰अयोः c. . .. . .॰आनाम् f. . .. . .
स॰ L.॰ए. . .॰आयाम् d॰एषु c & f. . .॰आसु c
a.These terminations are for nouns and adjectives. There are other sets of terms. (ch. 4) for numerals and pronouns, including pronominal adjs. (4.16:) – many of which also end in अ or आ.
b.The f. voc. of अम्बा “mother” is the simpler अम्ब (the first word a child learns), instead of अम्बे (perhaps an indication that Saṃskṛta was once a natural language of the people).
c.These terms. are based on the common pronominal terms. (4.4:). M.n. pl. inst. ॰ऐः term. is from m.n. of the pronoun तद् (4.6:), and f. pl nom. voc. ॰आः is from fem. of तद् (4.6:).
d.These terms. are based on the common feminine terms. ऐ आस् आम् taken by feminine stems in आ इ ई उ ऊ, including pronominal stems (4.4:).
e.The lengthening of the अ and insertion of न् before the n. pl. nom. voc. acc. इ has affinity with the strong changeable stems in neuter (3.2.b:).
f.Dental न् of ॰आनाम् is susceptible to 2.92:, e.g., रामानाम्. Loc. षु from सु because of preceding vowel (2.101:).

3.29: Vowel Stems – ॰इ aहरि (ha̍r-i) m. “Lord Hari (Viṣṇu)
वारि (vā̍r-i) n. “water” - similarly n. ॰तृ (3.40:)
मति (mat-i) f. “thought”
(exceptions to the normal case terms. are in bold)
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
प्र॰ N.॰इः॰इ a. . .॰ई॰इनी. . .॰अयः॰ईनि. . .
सं॰ V.॰ए. . .
द्वि॰ A.॰इम्. . .॰ईन्॰ईः
तृ॰ I.॰इना. . .॰या॰इभ्याम्. . .. . .॰इभिः. . .. . .
च॰ D.॰अये॰इने a॰यै or ॰अये b॰इभ्यः. . .. . .
प॰ Ab.॰एः॰इनः a॰याः or ॰एः b
ष॰ G.॰योः॰इनोः a. . .॰ईनाम्. . .. . .
स॰ L.॰औ c॰इनि a॰याम् or ॰औः b॰इषु. . .. . .
a.Neuter adjectives (not substantives) in ॰इ, ॰उ or ॰तृ (3.40:) may be declined like masculine throughout (ex. sg. du. pl. nom. voc. acc.) (–neut. ॰तृ declining like masc. ॰तृ, not masc. ॰इ). But neuter adjs. and subst. in ॰इ, ॰उ also may be declined like masculine in rare sg. voc.
b.Feminine adjectives and substantives in ॰इ or ॰उ may be declined like masculine in sg. dat. abl./gen. loc., in place of the common feminine terminations.
c.Apart from this vṛddhi of उ instead of इ, stems in ॰इ are declined similar to stems in उ.

3.30: Vowel Stems – ॰उगुरु (gur-) m. “teacher”
मधु (ma̍dh-u) n. “honey”
धेनु (dhen-) f. “cow” c
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
पुं॰ m.नपुं॰ n.
स्त्री॰ f.
प्र॰ N.॰उः॰उ a. . .॰ऊ॰उनी. . .॰अवः॰ऊनि. . .
सं॰ V.॰ओ. . .
द्वि॰ A.॰उम्. . .॰ऊन्॰ऊः
तृ॰ I.॰उना. . .॰वा॰उभ्याम्. . .. . .॰उभिः. . .. . .
च॰ D.॰अवे॰उने a॰वै or ॰अवे b॰उभ्यः. . .. . .
प॰ Ab.॰ओः॰उनः a॰वाः or ॰ओः b
ष॰ G.॰वोः॰उनोः a. . .॰ऊनाम्. . .. . .
स॰ L.॰औ c॰उनि a॰वाम् or ॰औः b॰उषु. . .. . .
a.See 3.29.a:.
b.See 3.29.b:.
c.Fem. adjs. in ॰उ denoting a quality, e.g., मृदु adj. “soft,” optionally add fem. suffix ॰ई (→॰वी 2.11:), declining like नदी.

3.31: There are several irregular declensions of certain common words in short ॰इ and ॰उ.

3.32: पति (pa̍t-i) m., when it means “husband” (not “lord”) and is not at the end of a cmpd., declines irregularly in sg. inst. dat. abl. gen. loc. as follows. Otherwise, like हरि.
m. पत्॰ए॰ Singular
पुं॰ m.
तृ॰ I.॰या
च॰ D.॰ये
प॰ Ab.॰युः
ष॰ G.
स॰ L.॰यौ

3.33: सखि (sa̍kh-i) m., “friend” when not at the end of a cmpd., declines irregularly like पति in sg. inst. dat. abl. gen. loc.
Also, like the changeable stems, it has a strong stem formed with vṛddhiसखाय् (sa̍kh-āy).
When at the end of a cmpd. it is irregular only in exhibiting this strong stem, the rest is like हरि.
Feminine is सखी, declined like नदी.
m. सख्॰ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.पुं॰ m.पुं॰ m.
प्र॰ N.॰आ॰आयौ॰आयः
सं॰ V.॰ए
द्वि॰ A.॰आयम्. . .
तृ॰ I.॰या. . .. . .
च॰ D.॰ये. . .
प॰ Ab.॰युः
ष॰ G.. . .. . .
स॰ L.॰यौ. . .

3.34: The neuter words अक्षि (a̍kṣ-i) “eye,” अस्थि (a̍sth-i) “bone,” दधि (da̍dh-i) “curd” and सक्थि (sa̍kth-i) “thigh” are irregular in the weak cases (vowel initial terms) for inst. through loc., in which they decline similar to the weak form of the three-form stem ॰अन् (3.20:), the इ of the stem being replaced with the below, e.g., sg. inst. अक्ष्णा a akṣ-ṇā. Shown are just the irregular forms.
ए॰ Singularद्वि॰ Dualब॰ Plural
नपुं॰ n.नपुं॰ n.नपुं॰ n.
तृ॰ I.॰ना. . .. . .
च॰ D.॰ने. . .
प॰ Ab.॰नः
ष॰ G.॰नोः॰नाम्
स॰ L.॰नि or ॰अनि (cf. 3.20.c:). . .
a.Rule 2.92:, न् will become ण् after अक्ष्, e.g., अक्ष्णा.

3.35: द्यु (dy-) m., “sky,” is irregular in that it takes vṛddhi in sg. nom. voc. and takes the normal case terminations in the rest, assuming the stem दिव् before an initial vowel term.
m. द्यु॰
/ दिव्॰
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.पुं॰ m.पुं॰ m.
प्र॰ N.द्यौःDual does not occur
in the literature
सं॰ V.
द्वि॰ A.दिवम्
तृ॰ I.दिवाद्युभिः
च॰ D.दिवेद्युभ्यः
प॰ Ab.दिवः
ष॰ G.दिवाम्
स॰ L.दिविद्युषु

3.36: Vowel Stems – ॰ई/॰ऊ
Feminine nouns and adjectives – including adj. cmpds. in any gender (mono/polysyllabic distinction pertaining only to the final cmpd. member only)
monosyll.धी (dh-ī̍) f. “thought”
भू (bh-) f. “earth”
polysyll.नदी (nad-ī̍) f. “river”
वधू (vadh-ū̍) f. “woman”
Stems in ॰ई and ॰ऊ are similarly declined, but differ between monosyllabic and polysyllabic stems, except in the pada terminations.
  • Monosyllabic stems before vowel terms. change ई/ऊ to इय्/उव् (2.67:).
  • Polysyllabic stems before vowel terms. change ई/ऊ to य्/व् (2.11:).
All take the normal case terminations, with a few exceptions in bold.
f. ध्॰/भ्॰
ए॰ Singularद्वि॰ Dualब॰ Plural
प्र॰ N.॰ईः a॰ऊः a॰ई c॰ऊः॰इयौ॰उवौ॰यौ॰वौ॰इयः॰उवः॰यः॰वः
सं॰ V.॰इ a॰उ a
द्वि॰ A.॰इयम्॰उवम्॰ईम् d॰ऊम् d॰ईः d॰ऊः d
तृ॰ I.॰इया॰उवा॰या॰वा॰ईभ्याम्॰ऊह्याम्. . .. . .॰ईभिः॰ऊभिः. . .. . .
च॰ D.॰इयै b
or ॰इये
॰उवै b
or ॰उवे
॰यै॰वै॰ईभ्यः॰ऊभ्यः. . .. . .
प॰ Ab.॰इयाः b
or ॰इयः
॰उवाः b
or ॰उवः
ष॰ G.॰इयोः॰उवोः॰योः॰वोः॰इयाम्॰उवाम्॰ईनाम्॰ऊनाम्
स॰ L.॰इयाम् b
or ॰इयि
॰उवाम् b
or ॰उवि
॰याम्॰वाम्॰ईषु॰ऊषु. . .. . .
a.The monosyllabic stems use the nom. form in sg. voc., instead of normal case term. Polysyllabic stems shorten the vowel in sg. voc.
b.The monosyllabic stems may optionally take the common feminine term. (but note the resulting identity between the pl. gen. and the optional sg. loc.), the polysyllabic must.
c.The polysyllabic ॰ई in sg. nom. drops the normal case term. स्, however it is retained in sg. nom. लक्ष्मीः “goddess Lakṣmī” and तन्त्रीः “string,” and optionally retained in तन्द्री a. “sloth.”
d.d. Polysyllabic sg. pl. acc. do not change ई/ऊ to य्/व्, but instead drop the vowel of normal case terms. अम् अस् to म् स्, respectively.

3.37: स्त्री (str-ī̍) f., “woman” is irregularly declined in that, although it takes the monosyllabic stem ॰इय् before vowel terms. (opt. in sg. pl. acc.), it otherwise takes the polysyllabic नदी terms., naturally taking only (not optionally) the three feminine terms. The pl. gen. form does not take ॰इय् before the inserted न्, and by sandhi (2.92:) converts it to ण्.
ए॰ Singularद्वि॰ Dualब॰ Plural
स्त्री॰ f.स्त्री॰ f.स्त्री॰ f.
प्र॰ N.स्त्रीस्त्रियौस्त्रियः
सं॰ V.स्त्रि
द्वि॰ A.स्त्रियम् or स्त्रीम्स्त्रियः or स्त्रीः
तृ॰ I.स्त्रियास्त्रीभ्याम्स्त्रीभिः
च॰ D.स्त्रियैस्त्रीभ्यः
प॰ Ab.स्त्रियाः
ष॰ G.स्त्रियोःस्त्रीणाम्
स॰ L.स्त्रियाम्स्त्रीषु

3.38: Vowel Stems – ॰ऋ
दातृ (dā̍t-) m. “giver”
पितृ (pit-ṛ̍) n. “father”
Agent nouns or Names of relations
(unusual forms in bold)
Mostly तृ suffix (6.26:), originally a cons. stem in ॰अर् (cognate to English -er), declines similar to three form stems in ॰अन् (3.20:).StrongMiddleWeak
॰आर् (vṛddhi) m. names of agents
॰अर् (guṇa) m.f. names of relations
॰ऋ॰र् / ॰ॠ
Ag. दात्॰
Rel. पित्॰
ए॰ Singularद्वि॰ Dualब॰ Plural
प्र॰ N.॰आ॰आरौअरौ॰आरः॰अरः
सं॰ V.॰अर्
द्वि॰ A.॰आरम्॰अरम्॰ॠन्॰ॠन् m., if f. ॰ॠः a
तृ॰ I.॰रा॰ऋभ्याम्॰ऋभिः
च॰ D.रेऋभ्यः
प॰ Ab.॰उर् (e.g. दातुः)
ष॰ G.॰रोः॰ॠणाम्
स॰ L.॰अरि॰ऋषु
a.Feminine relations differ from masculine only in pl acc. For feminine agents and Neuters see 3.40:

3.39: Irregular forms of ॰ऋ.
नप्तृ (na̍pt-) m. “grandson”
भर्तृ (bhar-tṛ̍) m. “husband”
Though relations, decline like agents.
स्वसृ (sva̍s-) f. “sister”Though relation, declines like agent. However, f. pl acc. remains स्वसॄः.
नृ (n-) m. “man”Declines like relations. However, pl. gen. opt. takes ॰ऋणाम्. (In Veda, sg. dat. नरे and sg. gen. नरः).
क्रोष्टृ (kroṣ-tṛ̍) m. “jackal” (lit. “yeller”)Takes ॰उ, instead of ॰ऋ, in middle form, ∴ pl. inst. क्रोष्टुभिः.

3.40: The suffix ॰तृ in other genders.
Neuter form of stems in ॰तृDecline like n. वारि (3.29:), replacing the vowel इ with ऋ, e.g., दातृ “giver” in n. sg. nom. is दातृ, sg. inst. is दातृणा. See also 3.29.a:, where it optionally declines like masc. ॰तृ in the inst. through loc. cases.
Feminine form of agent stems in ॰तृAdd feminine suffix ई (cf. 3.9:), e.g., दातृ “giver” in f. sg. nom. is दात्री (2.11:) and declines like नदी (3.36:).

3.41: Vowel Stems – ॰ऐ/॰ओ/॰औ. Only four common words end in diphthongs.
All are single syllables and take the normal case terminations, except sg. voc. declines like sg. nom.
रै (r-ai) m. “wealth”Before cons. रा॰, before vow. राय्॰ (2.71:).
नौ (n-au̍) f. “ship”Before cons. नौ॰, before vow. नाव्॰ (2.71:). Declines like रै.
गो (g-) m.f. “bull, cow”Before cons. गौ॰, before vow. गव्॰ (2.71:).
In strong cases b/4 cons. गौ॰, b/4 vow. गाव्॰ – गा॰ in sg. pl. acc.
Sg. abl. gen. गवस् contracts to गोस्.
द्यो (dy-) f. “sky”द्यो (dy-o) declines like गो (g-o), i.e., द्यौः (dy-auḥ), द्यावौ, द्यावः, etc.
ए॰ Singularद्वि॰ Dualब॰ Plural
प्र॰ N.राःनौःगौःद्यौःरायौनावौगावौद्यावौरायःनावःगावःद्यावः
सं॰ V.
द्वि॰ A.रायम्नावम्गाम्द्याम्गाःद्याः
तृ॰ I.रायानावागवाद्यवाराभ्याम्नौभ्याम्गोभ्याम्द्योभ्याम्राभिःनौभिःगोभिःद्योभिः
च॰ D.रायेनावेगवेद्यवेराभ्यःनौभ्यःगोभ्यःद्योभ्यः
प॰ Ab.रायःनावःगोःद्योः
ष॰ G.रायोःनावोःगवोःद्यवोःरायाम्नावाम्गवाम्द्यवाम्
स॰ L.रायिनाविगविद्यविरासुनौषुगोषुद्योषु

Go to Top ('Home' key on many Windows browsers)

4.1 4.2–3 4.4–5 4.6–7 4.8–10 4.11–3 4.14–5
4.16 4.17–22 4.23–25 4.26 4.27–30

4.1: Pronouns. Pronouns function as a substitue for other nouns, denoting a person or thing previously specified or understood in the context. Like in English, they can be classified as personal, demonstrative, relative, and interrogative. They are declined in संस्कृत differently from other nouns and adjectives.
Additionally, there are reflexive, possesive, and adjectival pronouns, where some decline like pronouns, others like regular nouns or adjectives, while a few are indeclinables. For the syntax of pronouns see 7.6:.

4.2 Personal pronouns. अहम् (1st person, “I”) and त्वम् (2nd person, “you”). The 3rd person (personal) pronoun, “he, she, it,” is not often expressed (7.6.B:), but, if expressed, it is represented by the demonstrative pronouns (4.6:-.10:).
As 1st member in cmpd.Three basic stem forms in declension
अहम्“I, we, us, mine”मद्अस्मद्आवअस्म
त्वम्“you, yours”त्वद्युष्मद्त्वयुवयुष्म
अहम् and त्वम् have no gender distinctions.
Both personal pronouns commonly are also represented by what are called defective (not declined in all cases), enclitic (unaccented, 4.3:) forms, i.e., मा, त्वा, etc., only in acc., dat., and gen. (the even numbered – although there are scattered instances, prior to Panini, of their use in all the oblique cases (all cases except nom. and voc.).
As with all pronouns, there is no vocative case and the sg. abl. and gen. differ. For just the personal pronouns, the pl. dat. and abl. also differ.
Those in bold below differ significantly from their matching stem (म, त्व, etc.)
ए॰ Singularद्वि॰ Dualब॰ Plural
म॰ stemत्व॰ stemआव॰ stemयुव॰ stemअस्म॰ stemयुष्म॰ stem
प्र॰ N.अहम्त्वम्आवाम्युवाम्वयम्यूयम्
द्वि॰ A.माम्मात्वाम्त्वानौवाम्अस्मान्नःयुष्मान्वः
तृ॰ I.मयात्वयाआवाभ्याम्युवाभ्याम्अस्माभिःयुष्माभिः
च॰ D.मह्यम्मेतुभ्यम्तेनौवाम्अस्मभ्यम्नःयुष्मभ्यम्वः
प॰ Ab.मद्त्वद्अस्मद्युष्मद्
ष॰ G.मममेतवतेआवयोःनौयुवयोःवाम्अस्माकम् aनःउष्माकम् aवः
स॰ L.मयित्वयिअस्मासुयुष्मासु
a.The pl gen. अस्माकम्, युष्माकम् are properly n. sg. possessive adjectives meaning “belonging to us/our, you/your,” respectively.

4.3: Enclitics because they are unaccented (and unemphasized), are not allowed to be placed:
1.At beginning of a sentence, subordinate clause, or pāda (1.23:).
2.After vocatives, which are unaccented, except as first word in sentence – in which case, since vocatives are treated as not part of the sentence, an enclitic could not follow, as it would then violate 4.3.1: above.
3.Before unaccented particles “and,” वा “or,” and (an exclamation), or the accented particle एव “only” – which all give emphasis to a preceding word.

4.4: Common Pronominal Terminations (except for the personal pronouns) (cf. 3.2:).
The case endings left blank tend to vary for the following pronouns, and for adjectives which decline like pronouns.
The initial vowel of these terminations replaces the pronominal stem’s final अ vowel, but otherwise is replaced by the stem’s vowel with the stem employing its long vowel form before long term. initial vowels.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.
प्र॰ N.आनिआस्
द्वि॰ A.अम्आम्आन्
तृ॰ I.एनयाआभ्याम्आभिस्
च॰ D.स्मैस्यैएभ्यस्आभ्यस्
प॰ Ab.स्मात्स्यास्
ष॰ G.स्ययोस्एषाम्आसाम्
स॰ L.स्मिन्स्याम्एषुआसु

4.5: Relative proximity may be contrasted or contextually expressed by demonstrative pronouns:
इदम् / एनद्“this” (sometimes, relatively, “that”)Near(सन्निकृष्टम्)
तद्“that”Out of sight(परोक्षम्)

4.6: Demonstrative Pronoun – तद् (ta̍-d) “that, those” (also the personal pronouns “he, him, his, it, its, she, her, hers, they, them” 4.2:). Has stem in त॰. This and the rest of the following pronouns as 1st member in cmpd. take the neuter sg. nom. form, e.g., तद्॰. The forms in bold are those for which no pronominal termination is common.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.
प्र॰ N.सःतद्सातौतेतेतानिताः
द्वि॰ A.तम्ताम्तान्
तृ॰ I.तेनतयाताभ्याम्तैःताभिः
च॰ D.तस्मैतस्यैतेभ्यःताभ्यः
प॰ Ab.तस्मात्तस्याः
ष॰ G.तस्यतयोःतेषाम्तासाम्
स॰ L.तस्मिन्तस्याम्तेषुतासु

4.7: Demonstrative Pronoun – एतद् (e-ta̍-d) “this, these” (“he,” etc.). Has stem in एत॰. Forms by compounding ए॰ to the pronoun stem त॰. Declines exactly like तद् above, e.g., m. n. f. sg. nom. एषः एतद् एषाः…एतेषु एतासु.

4.8: Demonstrative Pronoun – इदम् (i-d-a̍m) “this/that, these/those” (“he,” etc.) refers to what immediately preceded, or is obvious, in the context. It has stems in अ॰, अन॰, इ॰, and इम॰, with the stem final अ or इ disappearing before आ॰ ए॰ initial terminations.
Both अ॰ and इ॰ stems for इदम्, showing only in singular forms below, are considered stem forms due to their appearance in the sense of “this” before adverbial suffixes common to other pronoun stems (6.19:).
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.
प्र॰ N.अयम्इदम्इयम्इमौइमेइमेइमानिइमाः
द्वि॰ A.इमम्इमाम्इमान्
तृ॰ I.अनेनअनयाआभ्याम्एभ्यःआभिः
च॰ D.अस्मैअस्यैएभ्यःआभ्यः
प॰ Ab.अस्मात्अस्याः
ष॰ G.अस्यअनयोःएषाम्आसाम्
स॰ L.अस्मिन्अस्याम्एषुआसु

4.9: Demonstrative Pronoun – अदस् (a-da̍-s) “that, those” (“he,” etc.). Has stems in अमु॰ अमू॰ अमी॰. The forms in bold either have no common pronominal terminations or, mostly because stem final ऊ replaces inital term. आ॰, differ from those teminations. Note that अस्मै is unusually the same for both masc. and fem. sg. nom., and that अमू is unusually the same in dual nom./acc. for all genders.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.
प्र॰ N.असौअदःअसौअमूअमीअमूनिअमूः
द्वि॰ A.अमुम्अमूम्अमून्
तृ॰ I.अमुनाअमुयाअमूभ्याम्अमीभिःअमूभिः
च॰ D.अमुष्मैअमुष्यैअमीभ्यःअमूभ्यः
प॰ Ab.अमुष्मात्मुष्याः
ष॰ G.अमुष्यअमुयोःअमीषाम्अमूषाम्
स॰ L.अमुष्मिन्अमुष्याम्अमीषुअमूषु

4.10: Demonstrative Pronoun – एनद् (ena-d) is unaccented enclitic (4.3), and defective “this, these” (“him,” etc.). Forms stem in एन॰. These pronoun forms are considered optional forms for एनद् and इदम् above, and are substituted for them especially in the sense of “the afore mentioned” (rendered by the personal pronouns “him, her, it,” etc.). Declines only in the following oblique (4.14:) cases acc. inst. gen. loc. as shown. The rest do not occur and are left blank.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.
द्वि॰ A.एनम्एनद्एनाम्एनौएनेएनान्एनानिएनाः
तृ॰ I.एनेनएनया
ष॰ G.एनयोः
स॰ L.

4.11: Relative Pronoun – यद् (ya̍-d) “which.” Has stem in य॰. Declines like तद् (त॰). Only top five cases shown; the rest are similar.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.
प्र॰ N.यःयद्यायौयेयेयानियाः
द्वि॰ A.यम्याम्यान्
तृ॰ I.येनययायाभ्याम्यैःयाभिः
च॰ D.यस्मैयस्यैयेभ्यःयाभ्यः
प॰ Ab.. . .. . .
. . .

4.12: Interrogative Pronoun – किम् (ki̍-m) “who, which, what?” Has stem in क॰. Declines like तद् (त॰), except in n. sg. nom. acc. किम्.
ए॰ Singularद्वि॰ Dualब॰ Plural
पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.स्त्री॰ f.
प्र॰ N.कःकिम्काकौकेकेकानिकाः
द्वि॰ A.कम्काम्कान्
तृ॰ I.केनकयाकाभ्याम्कैःकाभिः
च॰ D.कस्मैकस्यैकेभ्यःकाभ्यः
प॰ Ab.. . .. . .
. . .

4.13: In word derivation (6.19: & .22:–.27:) the interrogative pronominal stems क॰ कि॰ कु॰ are used.
In compound (6.29:) the stem किम्॰ or sometimes कु॰ is used.
4.14: Reflexive pronouns, even though in the singular, may express any person in any number, e.g., “myself,” “himself,” “yourselves.” When used as such, if and when declined, they are in the singular oblique cases (all cases except nom. and voc.).
Words that can take on the sense of a reflexive pronounReflexive prounoun usage
आत्मन् “self”Masculine noun in ॰मन् (3.22:)In singular oblique cases can be used as a reflexive pronoun.
स्व m.n.
स्वा f. “own”
Possessive adj. declined per 4.16.3:
स्वयम् “itself”Indeclinable. When not used as a reflexive, it means “spontaneously.”As reflexive, can have the sense of nom., inst., or sometimes gen.

4.15: Possessive pronouns are adjectives with the sense स्व “(one’s) own.” Decline per अ/आ (3.28:).
निज a.“inborn, native,” can be used in the sense of “(one’s) own.”
With stems of
personal pronouns
and with तद्
मद्॰ stem→ मदीय “my” sg.अस्मद्॰ stem→ अस्मदीय “our” pl.
त्वद्॰ stem→ त्वदीय “your” sg.युष्मद्॰ stem→ युष्मदीय “your” pl.
तद्॰→ तदीय “his, hers, its, theirs”
Withभवत्॰ (3.13.b:)→ भवदीय “your” (in respectful address)
With genitivesमम॰→ मामक “my”तव→ तावक “your”
Withभवत्॰→ भावत्क “your” (in respectful address)

4.16: Pronominal adjectives, some of which are derived from pronominal stems, follow pronominal declensions a त॰ (सः तद् सा, etc.) in total or in part, e.g., m. n. f. sg. nom. अन्यः अन्यद् अन्या.
Unlike pronouns, they may take voc., which differs from nom. only in the sg. (as adj. stems ending in ॰अ/॰आ 3.28:) – m. n. voc. सर्व (॰—) f. सर्वे (॰ए), except as noted.
1.Follow pron. decl. (त॰), thus taking ॰द् in n. sg. nom. voc. acc.
n. sg. nom. voc. acc.
एकतम“one (of many)”एकतमद्
कतम“which of many?”कतमद्
कतर“which of two?”कतरद्
2.Follow pron. decl. (त॰), taking ॰म् (like in इदम्) in n. sg. nom. acc.
n. sg. nom. acc. (voc. as in 4.16:)
उभय“both” (sg. pl. only) bउभयम्
सर्व“every, all”सर्वम्
विश्व“every, all”विश्वम्
3.Follow pron. decl. (त॰), like सर्व taking ॰म् in n. sg. nom. acc., and also may follow nominal decl. ॰अ/॰आ (3.28:) in m. n. sg. abl. loc. and in m. pl nom. (thus making their decl. identical to ॰अ/॰आ throughout in m./n./f. sg./du./pl. nom./voc./acc.).
n. sg.
nom. acc.
m. n. sg. loc.
अधर“inferior, western”अधरम्अधरस्मिन्/ अधरे
अन्तर“inner”अन्तरम्अन्तरस्मिन्/ अन्तरे
अपर“other, inferior”अपरम्अपरस्मिन्/ अपरे
अवर“posterior, western”अवरम्अवरस्मिन्/ अवरे
उत्तर“subsequent, northern”उत्तरम्उत्तरस्मिन्/ उत्तरे
दक्षिण“southern”दक्षिणम्दक्षिणस्मिन्/ दक्षिणे
पर“subsequent, other”परम्परस्मिन्/ परे
पूर्व“prior, eastern”पूर्वम्पूर्वस्मिन्/ पूर्वे
स्व“own”स्वम्स्वस्मिन्/ स्वे
4.Decline like ordinary adjs., except may follow pron. decl. (त॰) in oblique cases (other than nom. & voc.) of the singular.
m. sg. loc.
द्वितीय“second”द्वितीये/ द्वितीयस्मिन्
तृतीय“third”तृतीये/ तृतीयस्मिन्
5.Decline like ordinary adjs., except may follow pron. decl. (त॰) m. pl. nom.
m. pl. nom.
अर्ध“half”अर्धाः/ अर्धे
अल्प“little”अल्पाः/ अल्पे
कतिपय“some”कतिपयाः/ कतिपये
चरम“last”चरमाः/ चरमे
द्वय/द्वितय“twofold” (4.30:)द्वयाः द्वितयाः/ द्वये द्वितये
प्रथम“first”प्रथमाः/ प्रथमे
and similar words in the stems ॰य and ॰तय (4.30:).
a.Any of these pronominal adjectives at the end of a bahuvrīhi (adjectival) compound (6.39:), e.g., अदृष्ट-पूर्व adj. “not seen before” (lit. “whose prior was not seen”), decline like ordinary adjectives.
b.For dual, use उभ॰ (declined in dual only) m. du. nom. acc. उभौ, n. f. du. nom. acc. उभे.

4.17: Pronominal stems + ॰दृश्, ॰दृश or ॰दृक्ष (“look” or “appearance”).
1.दृश् – Feminine is the same as masc. & neut. Decline like दिश् (3.6:)
तादृश्m.f.n. “like that”, “such”कीदृश्m.f.n. “like what?”
यादृश्m.f.n. “like which”मादृश्m.f.n. “like me”
ईदृश्m.f.n. “like this”त्वादृश्m.f.n. “like you”
2.दृश – Feminine is formed with . Decline like राम (3.28:), नदी (3.36:)
तादृशm.n. “like that”तादृशीf.
यादृशm.n. “like which”यादृशीf.
ईदृशm.n. “like this”ईदृशीf.
कीदृशm.n. “like what?”कीदृशीf.
मादृशm.n. “like me”मादृशीf.
त्वादृशm.n. “like you”त्वादृशीf.
3.दृश – Feminine is formed with . Decline like राम/सीता (3.28:)
तादृक्षm.n. “like that”तादृक्षाf.
यादृक्षm.n. “like which”यादृक्षाf.
ईदृक्षm.n. “like this”ईदृक्षाf.
कीदृक्षm.n. “like what?”कीदृक्षाf.

4.18: Pronominal stems + ॰वत् or ॰यत् (“much”), decline like ॰वत् (3.13:), the fem. forms with ई.
तावात्m.n. “that much”तावतीf.
एतावात्m.n. “this much”एतावतीf.
यावात्m.n. “as much”यावतीf.
इयत्m.n. “this much”इयतीf.
कीयत्m.n. “how much?”कियतीf.
4.19: Pronominal stems + ॰ति, decline only in pl. oblique cases like stems in ॰इ (3.29:).
ततिm.f.n. “that many”कतिm.f.n. “how many?”
यतिm.f.n. “as many”
4.20: Pronominal stem + ॰चिद्, ॰चन or ॰अपि make indefinite pronouns “someone, some.” With negative particle they means “no one, none.” The pron. stem can be declined in any gender, number or case before adding the particles (चिद्, etc.).
कोऽपि (2.32:)m.किमपिn.कापिf.
Similarly some adverbs are made into indefinite adverbs.
कदा “when?”→ कदाचिद् / कदाचन / कदापि “sometime or other,” “once,” with “never”
क्व “where?”→ क्वचिद् / क्वचन / क्वापि “somewhere or other,” with “nowhere”

4.21: Relative pronoun preceding an interrogative pronoun makes it indefinite, e.g.,
यः कःm. “whosoever”यः कश्चिद्m. “whoever”
यस्य कस्यm.n. “of whomever”
4.22: Relative pronoun when doubled takes on a distributive meaning (7.6.B.2.e:), e.g.,
यो यःm. “whoever,” “whatever in each case”


4.23: Cardinals – the numerals 0, 1, 2, etc. (‘. . .’ at 61 onward, only showing where optional forms or certain sandhis occur).
0 ०शून्यn.10 १०दशन्a.n.20 २०विंशति f.30 ३०त्रिंशत् f.40 ४०चत्वारिंशत् f.50 ५०पञ्चाशत् f.
1 १एकa.mnf.a11 ११एकादशन्a.n.b21 २१एक॰31 ३१एक॰41 ४१एक॰51 ५१एक॰
2 २द्व
in cmpd.
12 १२द्वादशन्a.n.c22 २२द्वा॰32 ३२द्वा॰42 ४२द्वा॰ or द्वि॰e52 ५२द्वा॰ or द्वि॰
3 ३त्रि
13 १३त्रयोदशन्a.n.d23 २३त्रयो॰33 ३३त्रयस्॰43 ४३त्रयश्॰ or त्रि॰53 ५३त्रयः॰ or त्रि॰
4 ४चतुर्
14 १४चतुर्दशन्a.n.24 २४चतुर्॰34 ३४चतुस्॰44 ४४चतुश्॰54 ५४चतुः॰
5 ५पञ्चन्a.n.15 १५पञ्चदशन्a.n.25 २५पञ्च॰35 ३५पञ्च॰45 ४५पञ्च॰55 ५५पञ्च॰
6 ६षष्a.n.16 १६षोडशन्a.n.d26 २६षड्॰36 ३६षट्॰46 ४६षट्॰56 ५६षट्॰
7 ७सप्तन्a.n.17 १७सप्तदशन्a.n.27 २७सप्त॰37 ३७सप्त॰47 ४७सप्त॰57 ५७सप्त॰
8 ८अष्टन् or अष्टौ a.n.18 १८अष्टादशन्a.n.28 २८अष्टा॰38 ३८अष्टा॰48 ४८अष्टा॰ or अष्ट॰58 ५८अष्टा॰ or अष्ट॰
9 ९नवन्a.n.19 १९नवदशन्a.n.29 २९नव॰39 ३९नव॰49 ४९नव॰59 ५९नव॰

60 षष्टि f.70 सप्तति f.80 अशीति f.90 नवति f.100 शतn.
. . .. . .. . .. . .101 एकशतn.
62 द्वा॰ or द्वि॰72 द्वा॰ or द्वि॰82 द्व्य्॰92 द्वा॰ or द्वि॰102 द्विशतn. dvi̍-śataf
63 त्रयः॰ or त्रि॰73 त्रयः॰ or त्रि॰83 त्र्य्॰93 त्रयः॰ or त्रि॰103 त्रिशतn. tri̍-śata
64 चतुः॰74 चतुः॰84 चतुर्॰94 चतुर्॰104 चतुःशतn.catu̍ḥ-śata
. . .. . .. . .. . .. . .
66 षट्॰76 षट्॰86 षड्॰96 षण्णवति
. . .. . .. . .. . .
68 अष्टा॰ or अष्ट॰78 अष्टा॰ or अष्ट॰88 अष्टा॰98 अष्टा॰ or अष्ट॰130 त्रिंशच्छतn. triṃśa̍c-chata

200 द्विशतn. dvi-śata̍ for द्वे शते1,000सहस्र n.or दश शतानि (lit. 10 100's)
300 त्रिशतn. tri-śata̍or त्रीणि शतानि10,000अयुत n.or दश सहस्रानि
400 चतुःशतn. catuḥ-śata̍or चत्वारि शतानि100,000लक्ष m.n.(= one lac)
500 पञ्चशतn. pañca-śata̍or पञ्चानि शतानि1,000,000नियुत n.or दश लक्षानि
10,000,000कोटि f.(= one crore) (=107)

1012महापद्मm.1017परार्धn. (the hightest number designation)

a.1-4 are adjectives taking all three genders; 5-19 are adjectives in neuter only; 20-99 are fem. substantives; and 100 onward are substantives of various genders (mostly neuter).
b.The form एका॰ is used only for 11, the rest – 21, 31, etc. take एक॰.
c.द्वा is an old dual form, द्वादशन् lit. means “two and ten.”
d.त्रयो॰ is from त्रि in m. pl. nom., त्रयः (4.25:). But षोडशन् shows phonetic changes similar to 2.100: & 2.109:, except there is no lost aspiration that would transfer to the following dental.
e.Alternate forms द्वि॰ त्रि॰ अष्ट॰ are used for 40’s-70’s, and 90’s, the 80’s have no alternates.
f.200, etc. differ from 102, etc. only by accent. शत śata̍ in compound is accented if 200, otherwise the added number (द्वि॰ dvi̍-) is accented for 102 (the same as for 11–99).

4.24: Alternate Cardinal Designations.
ऊन॰ a. “less”ऊनविंशतिः “twenty less (an understood one)” = 19
त्र्यूनत्रिंशत् “thirty less three (त्रि)”= 27
॰अधिक a. “plus”एकाधिकं शतम् “hundred plus one” = 101
द्व्यधिकशतम् = 102
Addition w/wo नव नवतिश्च or
नवतिर् नव “ninety and nine” = 99
Multiplication by
prefixed cardinal
त्रिदशन् tri-daśn “three times ten” = 30, the accent moving to the 2nd member of cmpd – but in unaccented texts this is in some cases ambiguous with the simple addition of the two members, e.g., अष्टशत may mean 108 (aṣṭa̍-śata) (4.23.f:) or 800 (aṣṭa-śat). An alternative to doubling or tripling a number is to use the dual or plural form of the number.

4.25: Declension of Cardinals
एक mn. एका f. adj. “one”Singular only, although its plural is used in an indefinite sense of “some,” “certain ones.” Declines like pronominal adj. सर्व (4.16.2:).
द्व mfn. adj. “two”Dual only, declines like stems in आ/आ (3.28:).
त्रि mn. तिसृ f. adj. “three”
चतुर् mn. चतुसृ f adj. “four”
Plural only, see declension below. Feminine चतुसृ declines like fem. तिसृ.
पञ्चन् “5” – नवदशन् “19” n. adj.Except in gen., those ending in न् decline like नामन् (3.22:) and षष् “six” decline like cerebral stems (3.6:), and are all plural neuter only, except sg. (in form only) in nom. voc. acc. See declensions below. The alternative form अष्टौ “8” is also as below.
These cardinals are employed like adjectives taking the same case as the numbered noun, although retain the neuter gender.
विंशति “20” and aboveSingular substantives, occasionally in plural as if used adjectively, e.g., शतानि “hundreds of __.” Their numbered noun in a sentence is in plural and takes the same case as this cardinal; or the numbered noun is in the genitive, as in the expression “a hundred of __ (gen.).”
These cardinals decline regularly like other nouns of their same gender ending in the same respective consonant (3.5:) or vowel.
“three”“four”“six”“five” – “nineteen”“eight”
त्रि tr-तिसृ tis-ṛ̍चतुर् cat-u̍rषष् ṣa̍-पञ्चन् pa̍ñc-an
॰दशन् -da̍ś-an
Also has older
अष्टौ aṣṭ-a̍u
पुं॰ m.नपुं॰ n.स्त्री॰ f.पुं॰ m.नपुं॰ n.नपुं॰ n.नपुं॰ n.नपुं॰ n.
प्र॰ N.॰अयः॰ईणि॰रः॰वारः॰वारि॰ट्॰अ॰औ
सं॰ V.
द्वि॰ A.॰ईन्॰उरः
तृ॰ I.॰इभिः॰ऋभिः॰उर्भिः॰ड्भिः॰अभिः॰आभिः
च॰ D.॰इभ्यः॰ऋभ्यः॰उर्भ्यः॰ड्भ्यः॰अभ्यः॰आभ्यः
प॰ Ab.
ष॰ G.॰अयाणाम्॰ऋणाम्॰उर्णाम्॰ण्णाम्॰आनाम्॰आनाम्
स॰ L.॰इषु॰ऋषु॰उर्षु॰ट्सु॰असु॰आसु


4.26: Ordinals are generally adjs. from cardinals having the sense of “__parts,” “__fold,” “containing__.” They are mostly derived from a small set of suffixes given below and declined in all genders like nominal stems in अ/आ.
Feminine is formed with ई, except 1st through 4th in आ.
Fractions may be distinguished from ordinals only by their accent shifting to the first syllable, e.g., तृतीय tṛtī̍ya “third” means “one-third” by moving the accent to first syllable tṛ̍tīya.
1stप्रथम or आद्यThese are not true ordinals, but have same meaning.
2ndद्वितीय॰ईय. For their declension see 4.16.4:.
4thचथुर्थ॰थ. Feminine in ई.
or तृतीय/ तुर्यOr ॰ईय/॰य. Feminine in आ.
These are likely an abbreviated form by dropping initial च.
6thषष्ठ॰थ (becomes ॰ठ after ष् – rule 2.88:).
11th–19thएकादश–नवदशSame form as cardinal, except dropping the final न् (hence declined different) and accent moving to last syllable.
20th, 99thविंश, त्रिंश, षष्ट, etc.Same as cardinal, shortening the cardinal until it ends in final अ (20-59) or replacing final इ with अ (60-99).
or त्रिंशत्तम–नवतितमOr simply by adding ॰तम to the full form of the cardinal.
100th, 1000thशततम, सहस्रतम॰तम.
101st, etc.एकशत, etc.The in-between ordinals (101st, 102nd, etc.) take same form as cardinal, with accent moving to last syllable. Hence, the pronunciation and declension of 200 (4.23.f:) and 102nd, etc. would be identical, apart from the cardinal being a neuter substantive while the ordinal an adjective in any gender.
or एकशततम, etc.Or simply by adding ॰तम.

Numerical Adverbs and Derivatives

4.27: Multiplicative adverbs
सकृत्“once” (lit. “one making”).
त्रिस्“three times.”
चतुस्“four times.”
॰कृत्वस्“__times,” e.g., पञ्चकृत्वः “five times.”
4.28: Adverbs of manner
॰धा“in __way(s)”, e.g., एकधा “in one way,” षोढा (cf. 2.100:, 2.109:) “in six ways,” etc.
4.29: Distributive adverbs
॰शस्“by __,” e.g., एकशः “singly,” द्विशः “by twos,” etc.
4.30: Aggregative adjectives and nouns
द्वयas adj. “twofold,” as a neut. noun “a pair.”
त्रयas adj. (fem. in ई) “threefold,” as a neut. noun “a triad” (group of three things).
॰तय“_fold,” e.g., त्रितय (same meanings as त्रय); चतुष्टय ca̍tuṣ-ṭaya (2.24:, .27:, .101:) as adj. “fourfold,” as a neut. noun “a tetrad,” etc.

Go to Top ('Home' key on many Windows browsers)

5.1–2 5.3 5.4–5 5.6 5.7–13 5.14–23
5.24 5.25–9 5.30–1 5.32 5.33–4 5.35
5.36–8 5.39 5.40 5.41 5.42 5.43

The Inflection of Verbal Roots by Means of Tense and Mood Terminations

5.1: The inflection of verbal roots is called conjugation. In conjugating संस्कृत verbal roots there are:
1.Three persons (पुरुष-s): third (“he, she, it”), second (“you”), first (“I”) (In the order recited they are called प्रथम “prior,” मध्यम “middle,” उत्तम “last” respectively, abbrev. as प्र॰ म॰ उ॰).
2.Three numbers (वचन-s): singular, dual, plural (like for nouns, abbrev. as ए॰ द्वि॰ ब॰).
3.Two categories of verbal terminations: परस्मैपद (ति तस् अन्ति…) and आत्मनेपद (ते ईते अन्ते…), abbreviated as पर॰ आ॰. These signify transitive (or active, lit. “for another”) and reflexive (or middle, lit. “for oneself”) terminations, respectively – adhered to less in the later literature. However, causal verb forms still nearly always respect this distinction.
4.Two categories of verbal terminations divided into two groups (see 7.21–7: for their syntax):
a. सार्वधातुक-लकार-s the four verbal base tenses and moods (also called the present system or the conjugational tenses and moods). Consisting of
  • Two tenses
  • Present tense, वर्तमाने लट्.
  • Imperfect (past) tense, अनद्यतन-भूते लङ्.
  • Two moods
  • Imperative mood, आज्ञायाम् लोट्.
  • Potential mood, विधि-लिङ्.
b. आर्धधातुक-लकार-s the six root tenses and moods (also called the non-conjugational tenses and moods). Consisting of
  • Four tenses
  • Perfect and Periphrastic Perfect (past) tense, परोक्ष-भूते लिट्.
  • Aorist (past) tense, अद्यतन-भूते or सामान्य-भूते लुङ्.
  • Simple future, सामान्य-भविष्यत्काले लृट्.
  • Periphrastic future, अनद्यतन-भविष्यत्काले लुट्.
  • Two moods
  • Benedictive mood, आशीर्-लिङ्.
  • Conditional mood, क्रियातिपत्तौ लृङ्.
5.Four derivative formations: Causative (णिजन्त), Desiderative (सन्नन्त), Intensive or Frequentive (यङन्त), and Denominative (नामधातु).
6.Three constructions (प्रयोग-s): Active (कर्तरि), Passive (कर्मणि) and Impersonal (भावे).

5.2: Roots. Roots are called धातु-s. They are the assumed basic unit of all verbs and participles, and most nominal stems. There is a listing of 2,200 roots, of which less than 500 are commonly used as verbs and participles. A root may be first made into a derivative form (5.39:-.43:), then that form is treated as a root taking one of the ten tenses or moods.

Terminations (5.5:–.6:). When a root, or a root derivative, takes one of the आर्धधातुक-लकार-s (the root tenses and moods) it undergoes a transformation unique to that tense or mood, and then takes a verbal termination. Except for the Perfect tense, which has its own set of terminations, the other tenses and moods generally take one or the other of the Present or Imperfect set of terminations from either the 1st or the 2nd conjugation. For this reason, in order to give them a generic name not tied to the Present or Imperfect tense, the Present tense terminations are also called the primary terminations (light shaded in 5.5:–6:), while those of the Imperfect are called the secondary terminations.

Verbal base (5.3:). When a root, but not a root derivative, takes one of the सार्वधातुक-लकार-s the four common verbal base tenses and moods) it is made into a stem – called a verbal base (अङ्ग) consisting of a root plus a sign. Most roots form their base in only one of ten different ways.

Conjugational groups. To help distinguish which ways these bases are formed, all the roots have been grouped into ten different classes (गण-s). Some roots have more than one base configuration, and therefore are placed in more than one class. These ten classes are arranged into two groups, the 1st conjugation (comprising the 1st, 4th, 6th and 10th class), otherwise known as the “गण-s or those classes whose base ends in अ, and the 2nd conjugation (comprising the 2nd, 3rd, 5th, 7th, and 9th class) or those classes whose base does not end in अ. The first group’s base is unchangeable when the terminations are applied; the second group’s bases are changeable into strong and weak forms, similar to nominal stems in declension.

Verbal constructions. The roots in the commonly used in active (कर्तरि) construction take one or the other of the two categories of terminations, परस्मैपद or आत्मनेपद, although परस्मैपद is the more common. Because of this, these roots are also called परस्मैपद or आत्मनेपद roots. Certain of these roots normally take only one of the categories of terminations, but may in composition take a termination from the other category due to meaning (5.1.3:), and occasionally due to meter or due to exceptional usage. Some roots take either of these sets of terminations and are therefore called उभयपद (lit. “word for both”) roots. The other two constructions (7.28:), in passive (कर्मणि) and in impersonal (भावे), force the root, no matter what its natural tendencies in active construction, to form a unique stem called a passive base which then takes the आत्मनेपद terminations of the 1st conjugation.

Verbal conjugation is a difficult part in learning the संस्कृत language. We will greatly simplify this process, however, because in this grammar we are just learning how to read the संस्कृत language. Our general assumption is that whoever wrote the text knows how to correctly construct the verb. We simply need to recognize which construction it is. Thankfully, these constructions follow a few recognizable patterns, so our job is not nearly as hard as it would seem. Your task, like it was in nominal declensions, is to grasp the patterns in the following charts. Take as much time as you wish to go through this – depending on whether you just want to get the general idea, memorize the most general features, or memorize most of it. So relax and let’s begin.

5.3: The Verbal Bases of the Ten Classes (गण-s) when forming the सार्वधाथुक-लकार-s.

1st Conjugation (unchangeable base)
1st173भू-आदिगुण of fin. vow. (w/sandhi 2.71:)
or light med. vow. e
endrootभवति bha̍v·a·ti (2.71:)
“he exists”
4th54दिव्‑आदिendrootदीव्यति dī̍v·ya·ti
(5.16.1:) “he plays”
6th35तुद्-आदिendsignतुदति tud··ti
“he strikes”
10th110चुर्-आदिवृद्धि of fin. vow. (w/sandhi 2.71:), गुण of light med. vow. ( usually lengthens).अय
endsignचोरयति cor·a̍ya·ti
“he steals”

2nd Conjugation (changeable base)
ClassQty.NameStrong FormWeak FormExampleg
2nd50अद्-आदिगुणrootterm.s. अत्ति a̍t·ti
w. अदन्ति ad·a̍nti
“he/they eat”
3rd11हु-आदिRedup.f & गुणrootRedup.term.जुहोति ju·h·ti
जह्वति ju·hv·a̍ti (2.11: & 5.6.c:) “he/they offer”
5th15सु-आदिनो hendsignनु hendterm.सुनोति su·no̍·ti
सुन्वन्ति su·nv·a̍nti
“he/they press”
7th14रुध्-आदि hb/4 root final cons.signन् hb/4 root final cons.term.रुणद्धि ru·ṇa̍·d·dhi
रुन्धन्ति ru·n·dh·a̍nti
“he/they oppose”
8th10तन्-आदिendsignendterm.तनोति tan··ti
तन्वन्ति tan·v·a̍nti
“he/they stretch”
9th23क्री-आदिना hendsignनी hb/4 cons.
term.क्रीणाति krī·ṇā̍·ti
“he buys”
न् hb/4 vow. terms.क्रीणन्ति krī··a̍nti
“they buy”
a.Number of commonly found roots within each class, as listed in Dhaturupamanjari.
b.Name given by grammarians. It takes the 1st entry in each class plus the word “etc.” (आदि).
c.This lists where the class sign is placed – either after the end of the root or just before the final consonant of the root.
d.The accent may fall on the root, the class sign, or the termination. When it falls on the root or class sign, there is often strengthening or lengthening of the vowel therein. Accent is the only difference between the 4th class verbal base and a passive stem (5.37:), and between a 10th class verbal base and some denominative stems (5.43:).
e.A light medial vowel refers to a metrically light syllable, where the vowel must be short and the following consonant within the root is a single (non-conjunct) consonant (see 1.22:).
f.Reduplication is an easy feature to recognize, but difficult to construct. See its rules in 5.7:.
g.The examples take the namesake root of each class in present tense 3rd person sg. and, for the 2nd conjugation roots to show both the strong & weak form, also the 3rd person pl. (5.6:).
h.5th, 7th and 9th class root penultimate nasal drops before adding the nasal class sign.

5.4: सार्वधातुक-लकार-s – The Four Verbal Base Tenses and Moods. The below terminations and their footnotes (5.5:–.6:) apply also to the आत्धधातुक-लकर-s and to derivative stems.

5.5: Terminations taken by the 1st Conjugation Classes of Roots. Final अ of verbal base joined to termination is shown in bold, followed by the actual termination. A mid-dot (·) is used to separate the visual components of these terminations, for pattern recognition.
ए॰ Sg.द्वि॰ Du.ब॰ Pl.ए॰ Sg.द्वि॰ Du.ब॰ Pl.
Primary / Present (वर्तमाने लट्)Verbal base + term.
प्र॰ 3rd अति ॰ति अतः ॰त·स् अन्ति ॰अन्ति अते ॰ते एते ॰ई·ते अन्ते ॰अन्ते b
म॰ 2nd असि ॰सि अथः ॰थ·स् अथ ॰थ असे ॰से एथे ॰ई·थे अध्वे ॰ध्वे
उ॰ 1st आमि ॰मि a आवः ॰व·स् आमः ॰म·स् ॰ए आवहे ॰व·हे आमहे ॰म·हे
Secondary / Imperfect (अनद्यतन-भूते लङ्)Augment  c + verbal base + term.
प्र॰ 3rd अत् ॰त् अताम् ॰ता·म् अन् ॰अन् अत ॰त एताम् ॰ई·ताम् अन्त ॰अन्त
म॰ 2nd अः ॰स् अतम् ॰त·म् अत ॰त अथाः ॰थास् एथाम् ॰ई·थाम् अध्वम् ॰ध्वम्
उ॰ 1st अम् ॰अम् आव ॰व आम ॰म ॰इ आवहि ॰व·हि आमहि ॰म·हि
Imperative (आज्ञायाम् लोट्)Verbal base + term.
प्र॰ 3rd अतु ॰तु अताम् ॰ता·म् अन्तु ॰अन्तु अताम् ॰ताम् एताम् ॰ई·ताम् अन्ताम् ॰अन्ताम्
म॰ 2nd  d अतम् ॰त·म् अत ॰त अस्व ॰स्व एथाम् ॰ई·थाम् अध्वम् ॰ध्वम्
उ॰ 1st आनि ॰आनि आव ॰आ·व आम ॰आ·म ॰ऐ आवहै ॰आ·व·है आमहै ॰आ·म·है
Potential (विधि लिङ्)Verbal base + term.
प्र॰ 3rd एत् ॰ई·त् एताम् ॰ई·ताम् एयुः ॰ईय्·उस् एत ॰ई·त एयाथाम् ॰ईया·ताम् एरन् ॰ईर्·अन्
म॰ 2nd एः ॰ई·स् एतम् ॰ई·तम् एत ॰ई·त एथाः ॰ई·थास् एयाथाम् ॰ईया·थाम् एध्वम् ॰ई·ध्वम्
उ॰ 1st एयम् ॰ईय्·अम् एव ॰ई·व एम ॰ई·म एय ॰ईय्·अ एवहि ॰ई·व·हि एमहि ॰ई·म·हि
a.Final अ of the verbal base or stem is lengthened before terms. beginning with म्॰ व्॰ (see 1st person).
b.Final अ of the verbal base or stem is dropped before terms. beginning with अ॰ आ॰ (including ए॰ ऐ॰, because their first component is अ॰).
c.An accented augment अ is put (after any prefix and) before the root, see 6.2:, in all forms of imperfect, aorist, and the conditional. This augment causes root initial छ्॰ to double (2.61:).This augment with an initial vowel takes वृद्धि, instead of गुण (2.20:). The augment is dropped in imperfect and aorist (both then taking an imperative meaning instead) after the negative particle मा “not” (7.22.4:).
d.Imperative पर॰ 2nd sg. term. is absent (like sg. voc. term. of ॰अ declension, e.g., राम गच्छ “Go, Rāma!”

5.6: Terminations taken by the 2nd Conjugation Classes of Roots. The few terminations that differ from those of the 1st conjugation classes of roots are in bold. The dark shaded inflections are strong, the rest are weak, and take strong/weak form bases per 5.3:.
ए॰ Sg.द्वि॰ Du.ब॰ Pl.ए॰ Sg.द्वि॰ Du.ब॰ Pl.
Primary / Present (वर्तमाने लट्)Verbal base + term.
प्र॰ 3rd ॰ति ॰तस् ॰अन्ति/अति c ॰ते ॰आते g ॰अते b
म॰ 2nd ॰सि ॰थस् ॰थ ॰से ॰आथे ॰ध्वे
उ॰ 1st ॰मि ॰वस् ॰मस् ॰ए ॰वहे ॰महे
Secondary / Imperfect (अनद्यतन-भूते लङ्)Augment + verbal base + term.
प्र॰ 3rd ॰त् ॰ताम् ॰अन्/अतु d ॰त ॰आताम् ॰अत b
म॰ 2nd ॰स् h ॰तम् ॰त ॰थास् ॰आथाम् ॰ध्वम्
उ॰ 1st ॰अम् ॰व ॰म ॰इ ॰वहि ॰महि
Imperative (आज्ञायाम् लोट्)Verbal base + term.
प्र॰ 3rd ॰तु ॰ताम् ॰अन्तु/अतु c ॰ताम् ॰आताम् ॰अताम् b
म॰ 2nd ॰धि/॰आन,॰हि/a ॰तम् ॰त ॰स्व ॰आथाम् ॰ध्वम्<
उ॰ 1st ॰आनि ॰आव ॰आम ॰ऐ ॰आवहै ॰आमहै
Potential (विधि लिङ्)Verbal base + term.
प्र॰ 3rd ॰यात् yā̍·t e ॰याताम् yā̍·tām ॰युस् y·u̍s e ॰ईत ॰ईयाथाम् ॰ईरन्
म॰ 2nd ॰यास् yā̍·s ॰यातम् yā̍·tam ॰यात yā̍·ta ॰ईथास् ॰ईयाथाम् ॰ईध्वम्
उ॰ 1st ॰यायम् yā̍·am f ॰याव yā̍·va ॰याम yā̍·ma ॰ईय ॰ईवहि ॰ईमहि
a.Imperative पर॰ 2nd sg. term. becomes:
  • ॰धि after stem final consonants.
    • 2nd class final lost aspiration will not be thrown backward, if possible (2.79:).
    • 9th class roots take ॰आन instead of ॰धि, without adding class sign नी.
  • ॰हि after stem final vowels, except:
    • 5th and 8th class roots with a single cons. before the final अ of their class sign drop the term. ॰हि.
b.The आ॰ 3rd pl. have no न् in the terms. of present, imperfect and imperative.
c.The 3rd class roots, and some other reduplicated stems, drop the न् in the present and imperative पर॰ 3nd pl. terms.
d.॰उस् is taken instead of ॰अन् in imperfect पर॰ 3nd pl.:
  • By 3rd class roots. Before which, a final simple liquid vowel takes गुण (ए ओ becoming अय् अव् per 2.13:–.14:).
  • By the roots विद् “know” and द्विष् “be displeased.”
e.पर॰ Potential terms. are essentially the same as या + 2nd conj. Imperfect terms. (cf. Benedictive 5.32:), in the 3rd pl. the या is reduced to य् before ॰उस्.
f.All accented weak terms. are accented on the first syllable of the termination.
g.Stem final vowels before all vowel initial terms. follow General Vowel Sandhi rules, except:
  • 2nd class weak stem final इ ई, उ ऊ before vowel initial terms. become इय् उव् (2.67:).
  • 3rd class weak stem final इ ई, उ ऊ of polysyllabic reduplicated stem after conj. cons. (e.g., ह्री “be ashamed”) and before vowel initial terms. become इय् उव् (2.67:).
h.A dental mute final of root may drop before imperfect 2nd sg. term. ॰स् instead of the term. per 2.24:, thus preserving the term. in the final verb form. Often, though, the preservation is by inserting a vowel between them, e.g., 5.19.4: & .19.6:.

5.7: Reduplication is taken by five different verbal formations, all take the General Rules.

5.8: General Rules of Reduplication. There are exceptions to both General and Specials Rules. The resulting first syllable of the reduplicated stem is below called the reduplicating syllable.
1.The first syllable ([cons.]+vowel) of root is reduplicated.बुध्बुबुध् bu-budh
2.Aspirated letters lose aspiration.भिद्बिभिद् bi-bhid
3.Gutturals become corresponding palatals, and the guttural ह् becomes ज्.कम्
कम् ca-kam
गम् ja-gam
जुहु ju-hu
4.Conjunct cons. are reduced to the 1st letter.स्यन्द्स्स्यन्द् sa-syand
a. But sibilant + hard cons. reduce to the latter.स्थास्था ta-sthā (5.8.6:)
5.Simple initial vowel is doubled.अस्स् ā-as
6.The following vowel in reduplicating syllable, if long, is shortened.भूबुभू bu-bhū
7.Non-final ए ऐ replicates with its 2nd component ,
ओ औ with (cf. 2.2:)
सिषेव् si-ṣev
लुलोक् lu-lok
8.Final diphthong vowel is treated as if आ, for this and all following formations, e.g., perf., (even non-redup.) ps., pp., etc. गै→ (गा) → गै ja-gai
5.9: Special Reduplication Rule for 3rd Class Roots (5.3:)
1.The vowels ऋ ॠ become .भृ
बिभृ bi-bhṛ
इयृ iy-ṛ (2.67:)
5.10: Special Reduplication Rules for Reduplicated Perfect (5.27:)
1.The vowels ऋ ॠ ऌ become .कृकृ ca-kṛ
2.Initial इ उ, if गुण or वृद्धि in final form, insert य् व् (2.67:).इष्→ st. 1st sg. इयेiy-e̍ṣ-a,
→ wk. 3rd pl. षुः i-iṣ-u̍ḥ
3.Roots with य अ, and liable to saṃprasāraṇa, replicate with इ उ.वच्→ st. 3rd sg. वाच u-vā̍c-a,
→ wk. 3rd pl. चुः u-uc-u̍ḥ
4.Roots with initial अ + conjunct cons. or initial ऋ + cons. replicate with आन्.अर्च्
आनर्च् ān-arc
आनृज् ān-ṛj
5.11: Special Reduplication Rules for Reduplicated Aorist (5.30.2.c:)
1.The vowels अ आ ऋ ॠ ऌ become .जन्→ (जिजन् see 5.11.2:)
2.Reduplicating vowel is lengthened if the syllable is metrically light. The resulting metrical syllables of a redup. aorist are thus light-heavy-light (U–U) 1.22:.जन्
→ 3rd sg. अजीजनत् a-jī̍-jan-at
→ 3rd sg. अजिग्रहत् a-ji̍-grah-at
5.12: Special Reduplication Rules for Desiderative (5.41:). The reduplicating syllable is accented.
1.The vowels अ आ ऋ (but not after labials which first change ऋ to ऊर्) replicate with .स्था
तिष्ठा ti̍-ṣthā
→ (भूर्) → बुभूर् bu̍-bhūr
2.Vowel initial roots replicate vowel-cons. and replace reduplicated vowel with .अश्
→ अशिश् a̍ś-iś
→ ईचिक्ष् ī̍c-ikṣ
→ एदिध् e̍d-idh
5.13: Special Reduplication Rules for Intensives (or Frequentives) (5.42:).
1.Reduplicating syllable takes गुण, will lengthen.लिह्
लेलिह् le-lih
तातप् -tap
2.Roots ending in अम् (sometimes other cons.) insert final cons., instead of lengthening अ.गम्
→ जङ्गम् ja-gam (2.55:)
→ चल्वल् cal-cal
3.Roots containing ऋ insert ई after reduplicating syllable.मृ
→ मरीमृ mar-ī-mṛ
→ चरीकृष् car-i-kṛṣ

5.14: Irregular Verbal Bases in सार्वधातुक-लकार-s – The Four Verbal Base Tenses and Moods.
for 1st Conjugational Classes of Roots
5.15: 1st Class (भू-आदि) – गुण of fin. vow. or light medial vow.; class sign is अ; accented root.
1.Lengthen vowel, instead of गुण.क्रम्
→ आचा
2.वृद्धि of vowel, instead of गुण.मृज्“cleanse”मार्जmā̍rj-a
3.Substitutes for अ.सद्“sink”सीद्sī̍d-a
4.Reduplicate with इ (originally 3rd class roots).घ्रा
5.Drop nasalदंश्

6.Substitute for root.गम्
ga̍cch-a (2.61:)
5.16: 4th Class (दिव्-आदि) – class sign is य; accented root.
1.Lengthen vowel.तम्
“be weary”
2.Drop nasalभ्रंश्
3.Saṃprasāraṇa (2.3:).व्यध्“pierce”विध्यvidh-ya
4.Substitutes for root.जन्“be born”जाjā̍-ya
5.17: 6th Class (तुद्-आदि) – class sign is अ; accented sign.
1.Insert nasal.कृत्
→ कृन्त
→ मुञ्च
→ लिम्प
→ लुम्प
→ विन्द
→ सिञ्च
2.Saṃprasāraṇa (2.3:).प्रछ्
3.Substitute for root.इष्

5.18: Irregular Verbal Bases in सार्वधातुक-लकार-s – The Four Verbal Base Tenses and Moods.
for 2nd Conjugational Classes of Roots
5.19: 2nd Class (अद्-आदि) – गुण with accented root in strong.
1. वृद्धि of roots ending with in strong b/4 cons. terms., and of मृज् “cleanse” in all strong forms. e.g., स्तु pr. 3rd sg. स्तौतिsta̍u·ti
मृज् pr. 3rd sg. मार्ष्टि
2. शी takes गुण in weak; inserts र् b/4 अ॰ terms. in pr. impf. impv. 3rd pl. शी “sleep” pr. 3rd sg. शेतेś·te
pr. 3rd pl. शेरते śe̍r·ate
3. वश् takes saṃprasāraṇa in weak. वश् “desire” pr. 3rd pl. उशन्तिuś·a̍nti
4. अस् drops initial in pot. and weak of pr. & impv., drops final स् b/4 pr. 2nd sg. ॰सि;
substitutes ए in impv. 2nd sg.; inserts ई b/4 स् त् of पर॰ impf. 2nd 3rd sg. (5.6.h:).
अस् “be” pot. 3rd sg. स्यात्s·yā̍t
pr. 3rd pl. सन्तिs·a̍nti
impv. 2nd sg. एधिe·dhi
impf. 3rd sg. आसीत्a̍-asī·t
5. हन् drops न् b/4 त्॰ थ्॰ terms. in weak;
weakens to घ् in pr. impf. impv. 3rd pl.;
weakens to in impv. 2nd sg.
हन् “destroy” pr. 2nd pl. हथha·tha̍
pr. 3rd pl. घ्नन्तिghn·a̍nti
impv. 2nd sg. जहिja·hi
6. Roots हन् “breathe,” जक्ष् “eat,” रुद् “weep,” श्वस् “breathe,” स्वप् “sleep”:
insert इ b/4 cons. terms. except य॰;
insert ई or b/4 स् त् of पर॰ impf. 2nd 3rd sg.
e.g., श्वस् pr. 3rd sg. श्वसितिśva̍si·ti
impf. 3rd sg. अश्चसीत्
 or अश्वसत्
7. ईड् “praise” and ईश् “rule” insert इ b/4 आ॰ cons. initial terms. in स्॰ ध्॰. e.g., ईश् pr. 2nd sg. ईशिषेī̍ś·i·ṣe
8. ब्रू inserts ई in strong b/4 cons. terms. ब्रू “speak” pr. 1st sg. ब्रवीमिbra̍v·ī·mi
9. अधि + √ “read” inserts य् b/4 vow. terms. in pr. & impf. (2.67:), with aug. अ becoming ऐय् (ai-y). अधि+ pr. 1st sg. अधीयेadhi·i·y·e̍
impf. 1st sg. अध्यैयिadhy·a̍i·y·i
10. चकास् “shine,” जक्ष् “eat,” जागृ (गुण b/4 उस्) “wake,” शास् “teach,” and दरिद्रा (final drops b/4 weak terms., inserts b/4 weak cons. terms.) “be poor” take 3rd class terms. अति उस् अतु pr. impf. impv. 3rd pl. (5.6.c:–.d:). e.g., जागृ pr. 3rd pl. जागरुःjā̍gr·ati
impf. 3rd pl. अजागरुःa̍·jāgar·uḥ
दरिद्रा st. pr. 3rd sg. दरिद्रातिdaridrā·ti
wk. pr. 3rd du. दरिद्रितःdaridr·i·taḥ
wk. pr. 3rd pl. दरिद्रतिdaridr·ati
11. शास् takes weak base शिष् b/4 cons. terms. (2.101:), ex. b/4 impv. 2nd sg. term. ॰धि where ॰स् drops (2.100:); takes अशात् as impf. 3rd sg. & opt. as 2nd sg. शास् “teach” pr. 3rd du. शिष्टःśiṣ·ṭa̍ḥ
impv. 2nd sg. शाधिśā[s]·dhi
impf. 3rd sg.
(opt. 2nd)
12. अद् inserts अ b/4 स् त् of पर्॰ impf. 2nd 3rd sg. अद् “eat” impf. 3rd sg. आदत्ā̍·ad·a·t
5.20: 3rd Class (हु-आदि) – Reduplicate; गुण with accented root in strong.
1. दा substitutes दद् in weak; impv. 2nd sg. देहि. दा “give” pr. 3rd sg. दत्तेdat·te̍
2. धा substitutes दध् in weak, becoming धत् b/4 त्॰ थ्॰ terms. (2.76: & .79:); impv. 2nd sg. धेहि. धा “place” pr. 3rd pl. दधतेda̍dh·ate
pr. 3rd sg. धत्तेdhat·te̍
3. मा “measure,” हा “move” in आ॰ take मिमी जिही as redup. base, drop the b/4 vowels. e.g., हा (आ॰) pr. 2nd sg. जिहीषेji·hī·ṣe
pr. 1st sg. जिहेji·h·e
4. हा “abandon” in पर॰ takes जही in weak; drops the b/4 vowels and य्॰ terms. हा (पर॰) pr. 3rd du. जहीतःja··taḥ
pr. 3rd sg. जहतिja·h·ati
5. हु takes impv. term. धि instead of हि. हु “sacrifice” impv. 2nd sg. जुहुधिju·hu·dhi
5.21: 5th Class (सु-आदि) – accented नो in strong; unaccented नु in weak.
1. Roots ending in vowels opt. drop उ of weak नु b/4 labial म्॰ व्॰ terms. e.g., सु “press out” pr. 1st du. सुन्वः
or सुनुवः
or su·nu·va̍ḥ
2. Roots ending in cons. insert व् b/4 vowel terms. (2.67:). e.g., शक् “be able” pr. 3rd pl. शक्नुवन्तिśak·nu·v·a̍nti
3. श्रु replaces रु with ऋ to become शृ. श्रु “hear” pr. 3rd sg. शृणोतिś·ṇo̍·ti
4. धू shortens to धु. धू “shake” pr. 3rd sg. धुनुतेdhu·nu·te̍
5.22: 8th Class (तन्-आदि) – accented ओ in strong; unaccented उ in weak.
1. कृ takes strong base in करो (kar-o);
takes weak base in कुरु (kur-u);
weak drops b/4 labial म्॰ व्॰ & य्॰ terms.
कृ “do” pr. 1st sg. करोमिkar·o̍·mi
pr. 3rd u. कुरुतःkur·u·ta̍ḥ
pr. 1st pl.कुर्मःkur·[u]·ma̍ḥ
2. Other roots opt. drop weak b/4 labial म्॰ व्॰ terms. तन् “stretch” pr. 1st du. तन्वः
or तनुवः
or tan·u·va̍ḥ
3. कृ with prefixes परि or सम्, and sometimes with उप, inserts स् between (6.12:). सम्+कृ “put together” pr. 3rd sg. संस्कुरुतेsaṃ·s·kur·u·te̍
5.23: 9th Class (क्री-आदि) – accented ना in strong; unaccented नी b/4 cons. or न् b/4 vow. terms. in weak.
1. मी “destroy,” ली “adhere,” धू “shake,” पू “purify,” लू “cut” shorten their vowel. e.g., धू pr. 1st sg. धुनामिdhu·nā̍·mi̍
2. ज्ञा “know,” ज्या “grow old,” ग्रह् “seize” shorten to जा, जि, गृह्. e.g., ज्ञा pr. 3rd sg. जानाति·nā̍·ti
3. Roots ending in ॠ shorten to b/4 cons. sign. e.g., गॄ pr. 3rd sg.गृणातिg·ṇā̍·ti

आर्धधातुक-लकार-s The Non-conjugational Tenses and Moods

5.24: For the आर्धधातुक-लकार-s (5.1.4.b:), the special verbal base (5.3:) of the 1st and 2nd conjugational classes of roots is not formed. Because of this, there is then no distinction here of the 10th classes of roots. Instead, the roots undergo general modifications unique to each one of these tenses and moods.

5.25: The Perfect tense (परोक्ष-भूते लिट्) expresses an action of the remote past, not witnessed by the speaker. As 2nd & 1st person are thus rarely found, then, in the following, the 3rd person is highlighted.

5.26: There are two types of Perfect tense.
Reduplicative Perfectfor primary (non-derivative) roots.Formed by replication (5.8: & .10:), and appending the Perf. terms. to the Perf. stem.
Periphrastic Perfectfor derivative stems (mostly
in ॰अय), and for a few roots.
Formed into abstract fem. noun in sg. acc. and adding the Perf. of अस् “be” भू “be” or कृ “do” (cf. 6.13:–.14:).

5.27: Reduplicative Perfect. The Perfect terminations, with dark shaded strong, are as follow. Accent is on root in strong; and on terminations in weak.
ए॰ Sg.2द्वि॰ Du.3ब॰ Pl.3ए॰ Sg.3द्वि॰ Du.3ब॰ Pl.3
प्र॰ 3rd अतुस् उस् आते इरे
म॰ 2nd (इ)थ 1 अथुस् (इ)से आथे (इ)ध्वे
उ॰ 1st (इ)व (इ)म (इ)वहे (इ)महे
1.Connecting vowel इ (shown in parenthesis above) is inserted before cons. terms., except:
For the rootsin उ:द्रु “run,” श्रु “hear,” स्तु “praise,” स्रु “flow”
in ऋ:कृ “do,” भृ “bear,” वृ “choose,” सृ “go”
Before पर॰ 2nd sg. , the is often dropped, especially:
For roots ending in ऋ, except ऋ “go” and वृ “cover”.
Optionally for roots ending in vowel other than ऋ.
Optionally for roots with a penultimate अ, e.g., पत् “fall” – पपत्थ or पपतिथ.
2.परस्मैपद Strong Stem formation
If root has3rd sg.2nd sg.1st sg.
a. Metrically light medial liquid vowelगुणगुणगुण
b. Medial अ followed by sing. cons.वृद्धिOpt. वृद्धि
c. Final vowelवृद्धिगुणवृद्धि or गुण
d. Final आ or diphthong plus term. (2nd sg. opt. acting as if weak, inserting इ, see below)आ·थ or इ·थ
(final vow. drops)
3.Weak Stem formation before vowel terms.
Final इ ई or preceded by sing. cons.→ य् or र्, respectively.
preceded by conj. cons.→ इय् (2.67:) or अर्, respectively.
Final उ ऊ or always become उव् (2.67:) or अर्, respectively.
Most roots in which is preceded and followed by a sing. cons. and the reduplicating syllable is the same as the root syllable, then contract the two into one syllable replacing अ with ए, e.g., पत् “fall” → अपत् pa-pa-t → weak stem पेत् pe-t.
This change also occurs in strong form 2nd sg., only when is inserted, e.g., पेति·थ or पपत्थ.
The medial अ roots – खन् “dig,” गम् “go,” घंस् “eat,” जन् “be born,” हन् “destroy” weaken by dropping the vowel, e.g., गम् → जगम् → weak stem जग्म्.
Roots in व्॰ followed by single cons. and √यज् “worship,” which replicate with उ इ further weaken the following radical व य by samprasāraṇa to उ इ, contracting to a long vowel, e.g., वच् “speak” → उवच् → weak stem ऊच्.
Roots with final आ or diphthong (5.8.8:) drop the vowel in the weak forms,
e.g., गै “sing” → strong जगा, weak जग् (cf. in 2nd sg. 5.27.2.d:).
4.Examples (in 3rd only, पर॰ sg. du. pl. & आ॰ sg. du. pl.)
भू“be” irreg.बभूवba-bhū̍-v-aबभूवतःba-bhū-v-a̍tuḥबभूवुःबभूवेबभूवातेबभूविरे

5.28: Irregular Forms of Perfect.
Contract two dissimilar syllables with
in weak
त्रस् “tremble” → तत्रस् त्रेस्
भज् “share” → बभज् भेज्
भ्रम् “wander” → बभ्रम् भ्रेम्
राज् “shine” → रराज् रेज्
No saṃprasāraṇa, and contract with ए in weak. यम् “reach” → ययम् येम्
वम् “vomit” → ववम् वेम्
वस् “wear” neither saṃprasāraṇa, nor contracts. वस् “wear” ववस्
Change radical consonant to guttural. चि “gather” चिकि
जि “conquer” जिगि
हन् “conquer” जघन्
हि “conquer” जिघि
अंश् “reach” reduplicates with आन् (5.10.4:), and drops radical nasal in weak. अंश् “reach” → strong आनंश्
→ weak आनश्
भू “be” reduplicates with ; and ऊ is unchanged, inserting व् b/4 vowels, वि (v-i) b/4 cons. (5.27.1:). भू “be” → 3rd pl. बभूवुः ba-bhū-v-u̍ḥ
→ 1st pl. बभूविम ba-bhū-v-i-ma̍
अह् “say” is defective, taking only 3rd sg. du. pl. & 2nd sg. du. (where ह् becomes त् before थ term.); often retains present tense meaning. अह् “say” → 3rd pl. आहुः a-ah-u̍ḥ
→ 2nd sg. आत्थ a-a̍t-tha
विद् “know” does not reduplicate; retains present meaning. विद् “know” → st. वेद्, wk. विद्;
वेद ve̍d-a “he/she/it knows”

5.29: Periphrastic Perfect is formed generally from a verbal stem that does not lend itself to reduplication, such as derivatives ending in अय/य (10th class stems, causatives and denominatives), the already reduplicated desideratives and intensives, and a few primary roots noted below.
Verbal stem made by गुण of fin. or short med. vowel is made into an abstract feminine noun in accusative with ॰आम् ā-m.+Reduplicated Perfect form of अस् “be” (3rd sg. आस, etc.) or, exceptionally, भू “be” (3rd sg. बभूव, etc.) or कृ “do” (3rd sg. चकार, etc.) – taking पर॰/आ॰ per the verbal stem.
E.g., दृश् “see” → दर्शय causal stem → Par. 3rd sg. दर्शयामास darś-aya-ām-a-a̍s-a
Par. 3rd pl. दर्शयामासुः darś-aya-ām-a-as-u̍ḥ
The original form of the Periphrastic Perf. was with transitive verb कृ, hence the acc. in formation even for the intransitive अस् and भू.
The primary roots taking Periphrastic Perfect, instead of Reduplicated Perfect are the following (shown in 3rd sg.). To show formation, the components are hypenated as if compounded words, but they should be written as a single word retaining the accent only of the abstract fem. noun:
Metrically heavy initial vowel आस्“sit”आसां-चक्रे
Already reduplicated चकास्“shine”चकासां-चकार
Optionally a few other roots, e.g. नी“lead”नयाम्-आस, or redup. perf. निनाय
भृ“bear”बिभरां-बभूव (3rd class roots redup. b/4 ॰आम्),
or redup. perf. बभार
ह्वे“call”ह्वयाम्-आस, or redup. perf. जुहव

5.30: Aorist (अद्यतन-भूते लुङ्) takes 7 Forms in two types (sibilant & simple). Grammarians differ in their numbering so will use their description (स-Aorist, etc.). All forms take the accented augment अ a̍ (5.5.c:) causing वृद्धि to initial vow. (2.20:), but is dropped when used in imperative sense with मा “not.” The terminations are based on the Imperfect (Secondary) terminations.
1.Sibilant Aorist inserts a sibilant between the root and before 2nd conj. Impf. terms. Before आ॰ 2nd pl. ॰ध्वम्, the final sibilant (॰स् or ॰ष्) of suffix converts ध् to ढ् (2.88:), then the sibilant drops.
a. स-Aorist is taken by a few roots ending in श् or ह् (both combining with स → क्ष kṣa 2.86:, 2.105:) and containing the vowels इ उ or ऋ.
Unmodified root+ ॰स
+ 2nd conj. Impf. terms. ex.1st conj. term.
आ॰ 3rd pl. अन्त.
b. स्-Aorist is taken by roots ending in cons. or in vowels other than आ.
पर॰: ऋद्धि of fin./med. vow.
आ॰: गुण of fin. vow.,
ex. ऋ.
+ ॰स्
+ 2nd conj. Impf. terms. ex.पर॰ 2nd sg. ईस्
पर॰ 3rd sg. ईत्
पर॰ 3rd pl. उस्.
c. इष्-Aorist, like the स्-Aorist above (5.30.1.a:), is taken by roots ending in cons. or in vowels other than आ.
पर॰: ऋद्धि of fin./med. vow.
आ॰: गुण of fin. vow., ex. ऋ.
+ ॰इष्
+ same terms. as स्-Aorist above, except suffix इष् drops before पर॰ 2nd 3rd sg. ईस् ईत् (likely replacements for iṣ-s and iṣ-t).
d. इष्-Aorist is mostly taken by six roots ending in ॰आ, inflected only in परस्मैपद.
Unmodified root+ ॰सिष्
+ same terms. as स्-Aorist above, except इष् of ॰इष् drops before पर॰ 2nd 3rd sg. ईस् ईत् (likely replacements for iṣ-s and iṣ-t).
2.Simple Aorist adds to the root w/wo connecting अ the 1st conj. Impf. terms.
a. अ-Aorist is taken by several roots, mostly पर॰, ending in cons.
Unmodified root+ ॰अ
+ 1st conj. Impf. terms. final अ treated as per 5.5.a:–.b:.
b. Root-Aorist is taken by a few roots ending in आ, and by the पर॰ root भू “be.”
Unmodified rootno
+ 1st conj. Impf. terms., ex. 2nd conj. पर॰ 3rd pl. उस्, before which a radical final आ is dropped..
c. Redulplcated-Aorist is taken by 10th class roots and the primary roots द्रु “run” श्रि “resort to,” and may give causal meaning to Aorist verb (e.g., नी “lead” → अनीनयत् a̍-nī-nay-a-t “caused to lead”).
Redup. root (5.8: & .11:)+ ॰अ
+ inflects like अ-Aorist (5.30.2.a:).

5.31 Examples of Aorist.
1.Sibilant Aorist
दिश् “point”→ पर॰ 3rdsg. अदिक्षत् a̍·dik·ṣa·t 2.86:,3rdpl. अदिक्षन् a̍·dik·ṣ[a]·an 5.5.b:
→ आ॰ 3rdsg. अदिक्षत a̍·dik·ṣa·ta,3rdpl. अदिक्षन्त a̍·dik·ṣ[a]·anta
दुह् “milk”→ दुह्+स → धुक्+ष 2.105:,.77:,.101:, 3rdsg. अधुक्षत् a̍·dhuk·a·t
“cut off”
→ पर॰ 3rdsg. अच्छैत्सीत् a̍·cchait·s·īt 2.77:,.42:,3rdpl. अच्छैत्सुः a̍·cchait·s·uḥ
→ आ॰ 3rdsg. अच्छित्त a̍·cchit·[s]·ta 2.98:,3rdpl. अच्छित्सत a̍·cchit·s·ata
श्रु “hear”→ पर॰ 3rdsg. अश्रौषीत् a̍·śrau··īt 2.101:,3rdpl. अश्रौषत a̍·śrau·ṣ·ata
Resulting sandhi with Aorist suffix स् before and after other cons.:
Before suffix स् final radical न् 2.93: and म् 2.104: become anusvāra, e.g.,
  • मन् “think” → अमंस्त a̍·ma·s·ta.
Before suffix स् final radical स् becomes dental त् 2.96:, e.g.,
  • वस् “dwell” → अवात्सीत् a̍·vāt·s·īt.
Suffix स् before soft dental ध्वम् cerebralizes latter to ढ्वम्, then drops 2.100:, e.g.,
  • अकृढ्वम् a̍·kṛ·[ḍhvam.
दह् “burn”+स् → धक्+ष् 2.105:,.77:,101:, e.g.,
  • अधाक्षीत् a̍·dhāk··īt.
दह्+स्+(त् or थ्) → दग्+[स्]+ध् → दग्ध् 2.76:,.78:,.98:, e.g.,
  • पर॰ 2nddu. अदाग्धम् a̍·dāg·dham, आ॰ 2ndsg. अदह्धाः a̍·dag·dhāḥ.
रुध् “hinder”+स् → रुद्+स् 2.76: → रुत्+स् 2.42: = रुस्, e.g.,
  • अरुत्सत a̍·rut·s·ata.
रुध्+स्+(त् or थ्) → रुद्+स्+ध् 2.76:,.78: → रुद्+[स्]+ध् 2.98: = रुद्ध्, e.g.,
  • 2nddu. अरौद्धम् a̍·raud·[sdham.
Irregular Forms of स्-Aorist
कृ “do” in आ॰ 3rd 2ndsg. drops the Aorist स् suffix, i.e.,
  • अकृत a̍·k·[s]·ta, अकृथाः a̍·kṛ·[s]·thāḥ.
Roots दा “give,” धा “place,” स्था “stand” weaken vowel to इ in आ॰, e.g.,
  • 3rdsg. अदित a̍·di·ta.
Roots दृश् “see,” सृज् “create,” स्पृश् “touch” take वृद्धि (आर्) transposed to रा in आ॰, e.g.,
  • 3rdsg. अद्राक्षीत् a̍·dk·ṣ·īt (2.86:)
वस् “stay” becomes वात् before the Aorist स् suffix, e.g.,
  • पर॰ 3rdsg. अवात्सीत् a̍·vāt·s·īt (2.96:).
पू “purify”→ पर॰ 3rdsg. अपावीत् a̍·pāv·[iṣ]·īt 2.71:,3rdpl. अपाविषुः a̍·pāv·iṣ·uḥ
→ आ॰ 3rdsg. अपविष्ट a̍·pav·iṣ·ṭa 2.98:,3rdpl. अपविषत a̍·pav·iṣ·ata
“be afraid”
→ पर॰ 3rdsg. अव्यथीत् a̍·vyath·[iṣ]·īt,3rdpl. अव्यथिषुः a̍·vyath·iṣ·uḥ
→ आ॰ 3rdsg. अव्यथिष्ट a̍·vyath·iṣ·ṭa 2.88:,3rdpl. अव्यथिषत a̍·vyath·iṣ·ata
Irregular Forms of इष्-Aorist
Medial vowel अ of the roots ending in र् and ल् (e.g. चर् and चल्), as well as the roots मद् “be exhilarated,” वद् “speak,” and व्रज् “move” takes वृद्धि in पर॰, e.g.,
  • 3rdsg. अमादीत् a̍-mād-[iṣ]-īt, 3rdpl. अमादिषुः a̍-mād-iṣ-uḥ.
या “purify”→ पर॰ 3rdsg. अयासीत् a̍·yā·[iṣ]·īt,3rdpl. अयासिषुः a̍·yā·siṣ·uḥ
2.Simple Aorist
→ पर॰ 3rdsg. अशुचत् a̍·śuc·a·t,3rdpl. अशुचन् a̍·śuc·[a]·an
→ पर॰ 1stsg. अशुचम् a̍·śuc·[a]·am,1stpl. अशुचाम a̍·śuc·ā·ma 5.5.a:
→ आ॰ 3rdsg. अशुचत a̍·śuc·a·ta,3rdpl. अशुचन्त a̍·śuc·[a]·anta
Irregular Forms of अ-Aorist
ख्या “tell” drops आ (cf. 5.30.2.b:), e.g.,
  • 3rdsg. अख्यत् a̍·khy[ā]·a·t, 3rdpl. अख्यन् a̍·khy[ā]·[a]·an.
दृश् “see” takes गुण, e.g.,
  • 3rdsg. अर्शत् a̍·darś·a·t.
अस् “throw” adds थ् to root, e.g.,
  • 3rdsg. आस्थत् a̍·asth·a·t (2.20:).
Roots पत् “fall,” वच् “speak” form contracted reduplicated Aorist, e.g.,
  • 3rdsg. अपप्तत् a̍·pa·pt·a·t, अवोचत् a̍·voc·a·t.
दा “give”→ पर॰ 3rdsg. अदात् a̍·dā·t,3rdpl. अदुः a̍·d[ā]·uḥ
Irregular Forms of Root-Aorist
भू “be” inserts व् before vowel terms., and 3rdpl. term. is 1st conj. अन्, e.g.,
  • पर॰ 1stsg. अभूवम् a̍·bhū·v·am, 3rdpl. अभूवन् a̍·bhū·v·an.
→ पर॰ 3rdsg. अमूचुचत् a̍··muc·a·t,3rdpl. अमूमुचन् a̍·mū·muc·[a]·an
→ आ॰ 3rdsg. अमूमुचत a̍·mū·muc·a·ta,3rdpl. अमूमुचन्त a̍·mū·muc·[a]·anta
Irregular Forms of Reduplicated-Aorist
व्यध् “pierce” takes saṃprasāraṇa (य to इ) to maintain the prevailing U–U (light-heavy-light) rhythm, e.g.,
  • 3rd sg. अवीविधत् a̍·vī·vidh·at.
Roots राध् “succeed” shortens radical syllable to produce the prevailing U–U rhythm, e.g.,
  • 3rd sg. अरीरधत् a̍·rī·radh·at.
However, roots दीप् “shine,” मील् “blink” do not maintain the prevailing rhythm (5.11.2:), keeping the reduplicating syllable short and the radical vowel long, e.g.,
  • 3rd sg. अदिदीपत् a̍·di·dīp·at, अमिमीलत् a̍·mi·mīl·at.

5.32: Benedictive mood (आशीर् लिङ्) inflects only in परस्मैपद. It is essentially formed by inserting स् between the या and the unique personal endings of the 2nd conj. Pot. terms. (5.6.e:).
Unmodified root+ या yā̍
+ स् s+ 2nd Conj. Imperfect terms. The 3rd pl. being यासुस्, also the स् drops before 2nd 3rd sg. terms. स् त् (cf. 5.30.1.c:), being then identical with the 2nd conj. Potential.
E.g., बुह् “awake” → पर॰ 3rd sg. बुध्यात् budh-yā̍-[s]-t, 3rd pl. बुध्यासुः budh-yā̍-s-uḥ

5.33: Simple Future (सामान्य-भविष्यत्-काले लृट्)
गुण of fin. vowel and metrically light medial.
10th class roots retain 1st conj. base with अय.
+ ()स्य
sya̍ or i-ṣya̍
+ 1st Conj. Present (Primary) terms.
Roots that take connecting vowel इ –
  • Most ending in cons. (only 100 do not), some optionally.
  • All ending in long (→ cons. ending अव्), or in ऋ ॠ (गुण → cons. ending अर्).
  • Only 12 roots ending in other vowels.
  • All derivatives (including 10th class), where अय drops final अ → अय् (cons.+इ).
कृ “do”→ पर॰ 3rdsg. करिष्यति kar·i·ya̍·ti,3rdpl. करिष्यन्ति kar·i·ṣy[a]·a̍nti (5.5.b:)
→ पर॰ 1stsg. करिष्यामि kar·i·ṣyā̍·mi,1stpl. करिष्यामः kar·i·ṣyā̍·māḥ (5.5.a:)
→ आ॰ 3rdsg. करिष्यते kar·i·ṣya̍·te,3rdpl. करिष्यन्ते kar·i·ṣy[a]·a̍nte
इ “do”→ पर॰ 3rdsg. एष्यति e·ṣya̍·ti,3rdpl. एष्यन्ति e·ṣy[a]·a̍nti
भू “do”→ पर॰ 3rdsg. भविष्यति bhav·i·ṣya̍·ti (2.71:)
Irregular Forms
ग्रह् “seize” (and opt. roots with final ॠ) inserts ई instead of connecting vowel इ, e.g.,
पर॰ 3rdsg. ग्रहीष्यति grah·ī·ṣya̍·ti (cf. 6.6: Irreg. ī-ta).
Most medial ऋ roots opt. transpose guṇa अर् to र b/4 स्य; a few necessarily, e.g.,
दृश् “see” and सृज् “emit” (cf. 5.31.1.b: Irreg.), e.g., दृश् → पर॰ 3rdsg. द्रक्ष्यति drak-ṣya̍-ti (2.86:).
Roots नश् “be lost” and मज्ज् “sink,” instead of गुण, optionally strengthen with a nasal before स्य, e.g.,
नङ्क्ष्यति nak-ṣya̍-ti or नशिष्यति naś-i-ṣya̍-ti, मङ्क्ष्यति mak-ṣya-ti (2.81:) or मज्जिष्यति majj-i-ṣya-ti.
वस् “dwell” changes its स् to त् before स्य, e.g.,
वत्स्यति vat-sya̍-ti (2.96:).

5.34: Periphrastic Future (or Second Future) (अनद्यतन-भविष्यत्-काले लुट्) in परस्मैपद only.
Root takes
गुण, like in
Simple Fut.
+ ()तृ in masc. nom.
(even if the agent is
fem. or neut.), i.e.,
()ता (i)tā̍ accented
+ in 3rd person: sg. du. pl. nom. of agent noun ॰तृ (3.38:).
in 2nd or 1st person: 2nd Conj. पर॰ Pr. tense of अस् “be” (5.19.4:), the agent noun ॰तृ remains in sg. even with du. and pl. अस्.
Roots that take connecting vowel इ –
Same roots as Simple Fut. स्य, except:
the roots गम् “go,” हन् “destroy,” and roots ending in do not insert इ.
भू “be”→ पर॰ 3rdsg. भविता bhav·i·tā̍,3rdpl. भवितारः bhav·i·tā̍r·aḥ
→ पर॰ 2ndsg. भवितासि bhav·i·tā̍·asi,2ndpl. भवितास्थ bhav·i·tā̍·stha
→ पर॰ 1stsg. भवितास्मि bhav·i·tā̍·asmi,1stpl. भवितास्मः bhav·i·tā̍·smaḥ
गम् “go”→ 3rdsg. गन्ता gan·tā̍ (2.55:), 2ndsg. गन्तासि gan·tā̍·asi, 1stsg. गन्तास्मि gan·tā̍·asmi
कृ “do”→ fut. stem करिष्य kar·i·ṣya̍, but Peri. Fut. पर॰ 1stsg. कर्तास्मि kar·tā̍·asmi

5.35: Conditional (क्रिया-अतिपत्तौ लृङ्) is rare. It has, as its formation reveals, a “past-future” meaning of “would have.”
Accented Aug. + Simple Fut. stem ()स्य+ 1st Conj. Imperfect (Secondary) terms.
भू “be”→ fut. भविष्य॰→ पर॰ 3rdsg. अभविस्यत् a̍·bhav·i·ṣya·t,3rdpl. अभविष्यन् a̍·bhav·i·ṣy[a]·an
→ आ॰ 3rdsg. अभविष्यत a̍·bhav·i·ṣya·ta,3rdpl. अभविष्यन्त a̍·bhav·i·ṣy[a]·anta
इ “go”→ fut. एष्य॰→ पर॰ 3rdsg. ऐष्यत् a̍i·ṣya·t (5.5.c:),3rdpl. ऐष्यन् a̍i·ṣy[a]·an

5.36: Passive formations. In passive (कर्मणि) and impersonal (भावे) construction, the verb takes the passive formation. The passive is formed generally in one of two ways:
a. For the six आर्धधातुक-लकार root tenses and moods, the आत्मनेपद terminations are able to assume a passive sense, as well as an active sense. For those forms that are only in परस्मैपद (e.g., Benedictive), they may also contextually assume the sense of the passive. The 3rd sg. Aorist, however, assumes a special form explained later.
b. For the four सार्वधातुक-लकार verbal base tenses and moods, instead of the roots taking the usual verbal base (5.3:) when forming the सार्वधातुक-लकार-s, a special passive base, no matter what the class of the root, is formed as follows.

5.37: Passive in सार्वधातुक-लकार-s.
The root is modified in this way:
1. Final or diphthong (5.8.8:) remains or
becomes ई (weakens), depending on the root.
+ ya̍
+ 1st Conj. आ॰ terms.
(regardless of the
class of the root) of
the appropriate
सार्वधातुक-लकार verbal
base tense or mood.
2. Final or is lengthened.
3. Final after single cons. becomes रि (2.70:), after conj. cons. becomes अर्.
4. Final becomes ईर् (2.68:), after labial becomes ऊर् (2.69:).
5. Nasal preceding a final cons. is dropped (weakens).
6. Saṃprasāraṇa, if the root is susceptible, is taken (weakens).
7. Derivative (and 10th class) stems in अय drop it (weakens), keeping the strong vowel of the derivative (including inserted प् of causal 5.40:).
भू “be”Present3rdsg. भूयतेbhū·ya̍·te,3rdpl. भूयन्तेbhū·y[a]·a̍nte
Imperfect3rdsg. अभूयतa̍·bhū·ya·ta,3rdpl. अभूयन्तa̍·bhū·y[a]·anta
Imperative3rdsg. भूयताम्bhū·ya̍·tām,3rdpl. भूयन्ताम्bhū·y[a]·a̍ntām
Potential3rdsg. भूयेतbhū·ya̍·ita,3rdpl. भूयेरन्bhū·ya̍·iran
धा “support”→ Present3rdsg. धीयतेdhī·ya̍·te,3rdpl. धीयन्तेdhī·y[a]·a̍nte
श्रु “hear”→ Present3rdsg. श्रूयतेśrū·ya̍·te,3rdpl. श्रूयन्तेśrū·y[a]·a̍nte
कृ “do”→ Present3rdsg. क्रीयतेk·ya̍·te,3rdpl. क्रीयन्तेk·y[a]·a̍nte
बन्ध् “bind”→ Present3rdsg. बध्यतेbadh·ya̍·te,3rdpl. बध्यन्तेbadh·y[a]·a̍nte
वच् “speak”→ Present3rdsg. उच्यतेuc·ya̍·te,3rdpl. उच्यन्तेuc·y[a]·a̍nte
कृ “do”→ Causal Pr.3rdsg. कार्यतेkār·[aya]·ya̍·te,3rdpl. कार्यन्तेkār·[aya]·y[a]·a̍nte
Irregular Forms
शास् “rule”→ Pr. 3rdsg. शास्यते śiṣ·ya̍·teor शिष्यते śiṣ·ya̍·te (weakens, cf. 5.19.11:)
खन् “rule”→ Pr. 3rdsg. खन्यते khan·ya̍·teor खायते khā·ya̍·te
तन् “rule”→ Pr. 3rdsg. तन्यते tan·ya̍·teor तायते tā·ya̍·te
वे “rule”→ Pr. 3rdsg. ऊयते ū·ya̍·te (weakens)
ह्वे “rule”→ Pr. 3rdsg. हूयते hū·ya̍·te (weakens)
5.38: Aorist 3rd sg. Passive takes a single special form (5.36.a:).
+ Strengthened root:+ the special 3rdsg.
आ॰ term. .
1. वृद्धि of final vowel.
2. Final आ inserts य् before the term. इ.
3. गुण of metrically light medial vowel, with अ lengthened.
4. Derivative अय drops (5.37.7:) before the term. इ.
श्रु “hear”→ आ॰ 3rdsg. अश्रावि a̍·śrāv·i
ज्ञा “know”→ आ॰ 3rdsg. अज्ञायि a̍·jñā·y·i
मुच् “release”→ आ॰ 3rdsg. अमोचि a̍·moc·i
पद् “go”→ आ॰ 3rdsg. अपादि a̍·pād·i
रुह् “sprout”→ causal रोपय॰ → आ॰ 3rdsg. अरोपि a̍·ro·p·[aya]·i
Irregular Forms
रभ् “seize” inserts a nasal, i.e.,
  • अरम्भि a̍·rambh·i
पॄ “fill” strengthens ॠ after labial to ऊर् (as per 5.37.4:), i.e.,
  • अपूरि a̍·pūr·i
Roots गम् “seize” रच् “fashion” वध् “slay” do not lengthen their अ, e.g.,
  • अगमि a̍·gam·i

Derivative Formations

5.39: Derivative formations. The prior verbal stem formations we have seen were for the most part unique for either the ten classes of roots in the सार्वधातुक-लकार-s, or for each one of the आर्धधातुक-लकार-s – though the 10th class root stem also appears in some of those latter tenses and moods, and thus arguably similar to the following derivative stems. Now the following secondary verbal stem formations, also called derivatives, differ in that the same stem formation with its own meaning attached can then be used in any of these tenses and moods, and even for forming nouns and adjectives.

5.40: Causatives.
Same strengthening as
10th class roots:
1. वृद्धि of final vow.
2. गुण of light medial
vow. (अ usually
+ अय a̍-ya
(Most roots
ending in आ
insert प् b/4
+ सार्वधातुक-लकार-s 1st Conj. पर॰ or आ॰ terms., regardless of natural tendencies of the root.
For आर्धधातुक-लकार-s, the strengthened ॰अय stem is treated as a root by the Paraphrastic Perfect, and in the other tenses and moods replacing अय with अयि, (the Benedictive पर॰ dropping the अय). However, the Reduplicated-Aorist itself can provide the causal sense for Aorist.
The passive accepts the strengthened causal stem, but drops the causal suffix अय.
Verbal nouns and adjectives may also be formed from the strengthened causal stem, with अय (usually अयि) or without.
कृ “do”→ कारय॰ “make” → cs. pr. 3rdsg. कारयति kār·a̍ya·ti
→ cs. ps. 3rdsg. कार्यते kār·[aya]·ya̍·te
सद् “sink”→ सादय॰ “destroy” → cs. pr. 3rdsg. सादयति sād·a̍ya·ti
→ cs. pot. 3rdsg. सादयेत् sād·a̍y[a]·et
दृश् “see”→ दर्शय॰ “show” → cs. impr. 2ndsg. दर्शय darś·a̍ya
→ cs. peri. perf. 3rdsg. दर्शयामास darś·a̍ya·ām·ās·a
स्था “stand”→ स्थापय॰ “stop” → cs. pr. 3rdsg. स्थापयति sthā·p·a̍ya·ti
→ cs. impv. 2ndsg. स्थापय sthā·p·a̍ya
धृ “hold”→ धारय॰ “sustain” → cs. pr. 3rdsg. धारयति dhār·a̍ya·ti
→ cs. impv. 2ndsg. धारय dhār·a̍ya
Irregular Forms
Roots ज्ञा “know,” ग्ला “languish,” म्ला “fade,” स्ना “wash” opt. shorten vowel b/4 पय, e.g.,
ज्ञापय॰ or ज्ञपय॰
Some roots not ending in आ also take प् before अय, and show final vowel irregularity, e.g.,जि “conquer”→ जपय॰ ja·p·a̍ya
अधि+इ “study”→ अध्यापय॰ adhy·ā·p·a̍ya “teach”
ऋ “go”→ अर्पय॰ ar·p·a̍ya
रुह् “grow” opt. takes प्, dropping the ह्, b/4 अय, i.e.,
रोहय॰ roh·a̍ya or रोपय॰ ro[h]·p·a̍ya “raise”
पा “drink” takes य्, instead of प्, b/4 अय, i.e.,
पायय॰ pā·y·a̍ya
Roots धू “shake,” प्री “love” strengthen with a nasal, i.e.,
धूनय॰ dhū·n·a̍ya, प्रीणय॰ prī··a̍ya
लभ् “take” strengthens medial vowel with an inserted nasal, i.e.,
लम्भय॰ lambh·a̍ya
दंश् “bite,” drops nasal in Present tense (दशति) 5.15.5:, but retains it in causal stem:
दंशय॰ daś·a̍ya
भी “fear” forms the optional stem:
भीषय॰ bhī·ṣ[a]·a̍ya (poss. den. 5.43: “intimidate”) or भायय॰ bhāy·a̍ya “raise”
हन् “destroy” forms the stem:
घातय॰ ghāt·[a]·a̍ya (poss. den. 5.43: “cause destruction”)

5.41: Desideratives.
Reduplicated root
(5.8: & .12:), the
reduplicating (1st)
syllable is accented
+ ()
sa or i-ṣa
+ 1st Conj. terms. for सार्वधातुक-लकार-s tenses and moods in पर॰ or आ॰, appropriate to the root.
Otherwise, the stem drops its final अ and is treated as a root for:
Periphrastic Perfect
इष् Aorist
Both Futures & Conditional, with (इ)
Adjective and noun formation
Before redup. and adding स, final इ उ are lengthened, final ऋ ॠ become ईर्, and after labials ऊर्.
The connecting vowel इ is added as in Simple Future स्य (5.33:), except:
Roots ending in short ऋ and long ऊ ॠ mostly do not take इ.
Roots ऋ “go,” दृ “hold,” पू “purify,” do take इ.
Adjectives are formed with , e.g.,
ज्ञा → जिज्ञासु ji̍·jñā·s[au “desiring to know” (5.8.4: & .12.1:).
Nouns are formed with feminine , e.g.,
ज्ञा → जिज्ञासा ji̍·jñā·s[aā “the desire to know”.
बुध् “know”→ des. pr.3rdsg. बुभुत्सति bu̍·bhut·sa·ti (2.76:–.77:),
3rdpl. बुभुत्सन्ति bu̍·bhut·s[a]·anti
जीव् “live”→ des. stem जिजीविष॰ ji̍·jīv·i·ṣa
आप् “obtain”→ des. stem ईप्स॰ [pip·sa (contracted from ip-ip, 5.12.2:)
कृ “do”→ des. stem चिकीर्ष॰ ci̍·kīr·ṣa (5.8.3:)
घस् “eat”→ des. stem जिघत्स॰ ji̍·ghat·sa (5.12.1: & 2.96:)
Irregular Forms
गम् “go” lengthens medial अ, i.e.,
जिगांस॰ ji̍·gāṃ·sa (2.54:)
मन् “think” lengthens medial अ and reduplicating vowel, i.e.,
मीमांस॰ mī̍·māṃ·sa (5.12.1:)
हन् “destroy” lengthens medial अ and reverts ह् to guttural, i.e.,
जिघांस॰ ji̍·ghāṃ·sa
Some roots with palatal, e.g., चि “gather” and जि “conquer” may revert to guttural, i.e.,
चिकीष॰ ci̍·kī·ṣa (or चिचीष॰) and जिगीष॰ ji̍·gī·ṣa
Roots ग्रह् “seize,” प्रछ् “ask,” स्वप् “sleep” take saṃprasāraṇa, i.e.,
जिघृक्ष॰ ji̍·ghṛk·ṣa (2.79: & .105:), पिपृच्छिष॰ pi̍·pcch·i·ṣa (2.61:), सुषुप्स॰ su̍·ṣup·sa
Seven roots contract the reduplicated root to one consonant after the reduplicating syllable:
दा “give”→ redup. दिदास॰→ des. stem दित्स॰di̍·t·sa
धा “place”→ redup. दिधास॰→ des. stem धित्स॰dhi̍·t·sa (2.76:–.77:)
मा “measure”→ redup. मिमास॰→ des. stem मित्स॰mi̍·t·sa
पद् “go”→ redup. पिपत्स॰→ des. stem पित्स॰pi̍·t·sa
रभ् “grasp”→ redup. रिरप्स॰→ des. stem रिप्स॰ri̍·p·sa
लभ् “take”→ redup. लिलप्स॰→ des. stem लिप्स॰li̍·p·sa
शक् “be able”→ redup. शिशक्ष॰→ des. stem शिक्ष॰śi̍·k·ṣa
Five roots conjugate as desideratives, but are generally said to have no obvious desiderative sense:
चित् “know”→ des. pr. 3rdsg. चिकित्सति“he/she cures”
गुप् “protect”→ des. pr. 3rdsg. जुगुप्सते“he/she depises/guards against”
तिज् “be sharp”→ des. pr. 3rdsg. तितिक्षते“he/she forbears”
→ f. तितिक्षा“forbearance”
बाध् “repel”→ des. pr. 3rdsg. बीभत्सते“he/she abhors”
मन् “think”→ des. pr. 3rdsg. मीमांसते“he/she considers/analyses”
→ f. मीमांसा“analysis”
5.42: Intensives (or Frequentives).
This rare form is taken by 60 monosyllabic roots, only a few beginning with a vowel.
First Form:
(mostly Vedic)
Reduplicated root
(5.8: & .13:)
+ ई opt. inserted
before cons. init.
strong terms.
+ 2nd Conj. पर॰ terms. and accenting like 3rd class roots w/गुण in strong form (5.3:), ex. accent on 1st syllable in strong and no गुण of medial root vowel before vowel ini. terms. or before the opt. ई.
Second Form:
Reduplicated root
(5.8: & .13:)
+ ya̍ accented
(a final vowel of root treated like before the passive ॰य)
+ 1st Conj. आ॰ terms. only.
For आर्धधातुक-लकार-s, the reduplicated ॰य stem is treated as a root, with connecting vowel as per the simple root’s conjugation. No instance of a Perfect is found, but likely Periphrastic Perfect would be used.
विद् “know”→ 1st Form s. वेवेद्॰, w. वेविद्॰ →पर॰ 3rdsg. वेवेत्ति ve̍·vet·ti or वेविदीति ve̍·vid·ī·ti
पर॰ 3rdpl. वेविदति ve·vid·a̍[n]ti (5.6.c:)
लिह् “lick”→ 2nd Form लेलिह्य॰ → आ॰ 3rdsg. लेलिह्यते le·lih·ya̍·te, 3rdpl. लेलिह्यन्ते le·lih·y[a]·a̍nte
Irregular Forms
गृ “awake” replicates with आ (as if from गर्) forming जागृ which seems to have replaced गृ as the 2nd class form of the root (5.19.10:).
Roots दह् “burn,” जभ् “snap at” replicate with nasal, i.e.,
दन्दहीति da̍-n-dah-ī-ti and दन्दह्यते da-n-dah-ya̍-te,
जञ्जभ्यते ja·ñ·jabh·ya̍·te
चर् “move” replicates with nasal and replaces radical vowel अ with ऊ, i.e.,
चञ्चूर्यते ca·ñ·cūr·ya̍·te
पद् “go” also inserts ई after the replicative nasal, i.e.,
पनीपद्यते pa·n·ī·pad·ya̍·te (cf. 5.13.3:)
द्रा “run” replicates as if from vowel ऋ and inserts a short इ, i.e.,
दरिद्राति da̍r·i·drā·ti (cf. 5.13.3:)

5.43: Denominatives are verbal stems derived from nouns and inflected like 1st conj. roots.
Modified noun:
1. Final often is lengthened, sometimes becomes .
2. Noun stems in अन् are treated like stems in .
3. Final इ उ are lengthened.
4. Final becomes री.
+ ya̍
+ 1st Conj. terms.
The आर्धधातुक-लकार-s are extremely rare.
Final vowel of noun is replaced by a, e.g.,
f. असूया → den. stem असूय॰ asūy[āa
Denominatives in short अ which have the causative accent (a̍[-ya]) are considered roots of the 10th class, e.g.,
मन्त्रय॰ (10th class मन्त्र्) “take counsel (मन्त्र),”
कथय॰ (10th class कथ्) “tell how (कथम्),” etc.
Denominatives express relations to the noun, depending on context, in these senses:
“be or act like,” “treat as,” “desire,” e.g.,
पीत्रीय॰“be like (treat like) a father (पितृ).”
राजाय॰“act like a king (राजन्).”
पुत्रीय॰“treat like (or desire) a son (पुत्र).”
“make,” and similar meanings, e.g.,
शब्दाय॰“make a noise (शब्द).”
दुःखाय॰“suffer misery (दुःख).”

***** Sanskrit Grammar *****