Derivation refers to the process of forming words on the basis of other words. This chapter will only treat of nominal derivation, that is, the process of forming nominals; the formation of verbs, including denominal verbs (those formed from an existing nominal base) and deverbal verbs (those formed from an existing verbal base), is treated in the chapter on verbal morphology.
Derivation in Sanskrit is straightforward. The process begins with a base aṅgam to which is added an suffix pratyayaḥ. The affix will sometimes induce additional changes in the base, either in terms of vowel gradation or accent. Pāṇini’s names for these affixes contain diacritics anubandhaḥ that represent the gradational and accentual properties of the resulting nominal form.
In a number of cases, the suffix is a null suffix: nothing is actually added to the base, although the base might be modified by the suffix in some way, but as a result of the addition of such a suffix, the base becomes a declinable nominal form.
Primary derivatives are those that are formed directly from verbal roots dhātuḥ. The suffixes that are added to verbal roots are called kr̥t, and hence primary derivatives are often called kr̥dantam “ending in a kr̥t suffix.”
In most cases, primary nominal derivatives have a meaning that is composed of (a) the meaning of the verbal root, (b) the particular thematic role that the suffix is associated with, representing a participant in the action of the verb, and (c) some additional component of meaning, related to time, aspect, and so on.
For adjectives formed from a verbal stem (participles), see the section on verbs. The following sections only discuss adjectives formed directed from a verbal root. They can also be called “participles,” because they are adjectival forms derived from verbal roots that often function as verbs in a sentence.
Pāṇini calls two sets of closely-related forms, those formed with the suffix Ktáḥ and those formed with the suffix KtávatU, by the common designation niṣṭhā́. These are both generally past verbal adjectives, in that they refer to an action that occurred prior to the time in which the statement is made. Hence they can be called “past participles” bhūtē kr̥dantāni.
The difference between them is the thematic roles (kārakāṇi, see below) to which they refer. Ktáḥ, which is often called the “past passive participle,” can refer to one the following three:
- the patient kárma of a transitive sakarmakaḥ verb;
- the agent kartŕ̥ of an intransitive akarmakaḥ verb (and sometimes of transitive verbs as well); and
- the verbal action itself bhāváḥ.
By contrast, KtávatU, which is often called the “past active participle,” can only refer to:
- the agent kartŕ̥, either of a transitive or intransitive verb.
Thus it is more accurate to think of Ktáḥ as an absolutive form, and KtávatU as a standard active form, of the past verbal adjective.
Many languages have nominative-accusative patterns of alignment, according to which the agent of a transitive verb, or an intransitive verb, takes one form (the nominative), and the patient of a transitive verb takes another form (the accusative). Some languages, however, have ergative-absolutive patterns of alignment, where the agent of an intransitive verb and the patient of a transitive verb take the same form (the absolutive) while the agent of a transitive verb takes another (the ergative).
Sanskrit, like many of the modern Indic languages such as Hindi, exhibits a split between the two patterns of alignment. Whenever finite verb forms are used, the alignment pattern is nominative-accusative. Whenever non-finite verb forms, such as participles, are used, the alignment pattern is ergative-absolutive. This more or less maps on to the distinction between past and non-past, since non-finite verb forms are very commonly used with reference to past time, and finite verb forms, by contrast, are typically used to refer to the present, or to hypothetical situations.
The form that Pāṇini calls Ktáḥ, that is, -tá-, is the primary form of what is often called the “past passive participle” or “p.p.p.” (Everything that is said in this section will apply equally to the form in -ná-, discussed below, which is a substitute for Ktáḥ after certain verbs.) This form is called the “past passive participle” for the following reasons:
- it is generally used in reference to the past bhūtḗ;
- it is commonly used in reference to the patient of a transitive verb (karmáṇi prayōgáḥ, see below), which is similar to passive constructions in English; and
- it is an adjectival form of a verb (a participle).
The initial K of Ktáḥ is an anubandhaḥ, that is, a marker that is not part of the form itself, but conveys information about its accentual and gradational properties. The K marker indicates that the form to which it is added does not undergo guṇáḥ, and therefore remains in the zero grade (or in some cases undergoes changes in order to reach the zero grade).
However, it can also be used of the agent of an intransitive verb. Pāṇini outlines the use of Ktáḥ as follows:
- It can refer to the patient kárma of a verb (i.e., of a transitive verb, since only such verbs have patients in the first place), according to Aṣṭādhyāyī 3.4.70. This is the most common use of the suffix. Some examples:
- dā + tá → dattá- “given”
- kr̥ + tá → kr̥tá- “done”
- dr̥ś + tá → dr̥ṣṭá- “seen”
- krī + tá → krītá- “bought”
- It can refer to the verbal action bhāváḥ in the case of intransitive verbs, also according to according to Aṣṭādhyāyī 3.4.70. Some examples:
- ās + tá → āsitá- “sitting down”
- śī + tá → śayitá- “lying down”
- It can refer to the agent kartŕ̥ in the case of intransitive verbs, according to Aṣṭādhyāyī 3.4.72. But in addition to intransitive verbs, Pāṇini lists a number of other verbs, which are either treated as transitive in the
Aṣṭādhyāyī or could possibly be transitive with certain verbal prefixes upasargā́ḥ, in order to ensure that Ktáḥ can be used to refer to the agent in the case of those verbs as well. The complete list is: verbs of movement gatyarthāḥ, śliṣ “embrace,” śī “lie down,” sthā “stand,” ās “sit,” vas “dwell,” jan “be born,” ruh “mount,” and jr̥̄ “grow old.” Some examples:
- rāmō vanaṁ gataḥ “Rāma went to the forest” (gam)
- sītā śayitā “Sītā laid down” (śīṄ)
- upāsitō guruṁ dēvadattaḥ “Dēvadatta served his teacher (ās)
- udayagirim adhirūḍhaḥ sūryaḥ “The sun ascended the eastern mountain” (ruh)
- It can be used to refer to the beginning of an action, with reference to either the agent or the patient, according to Aṣṭādhyāyī 3.4.71.
- prabhukta ōdanaṁ dēvadattaḥ “Devadatta has started to eat his rice.” (bhuj)
- prabhuktam ōdanaṁ dēvadattēna id.
See Speijer §360.
It is also worth noting that Ktáḥ generally refers to an action in the past, but in several cases, it is more salient that the consequences of that action have continued into the present, and hence the form refers to the present. For example:
aham iha sthitaḥ
“I have stood here” → “I am (standing) here” (sthā)
cf. Greek statives like ἕστηκα “I am standing”
- sō mr̥taḥ “He has died” → “He is dead” (mr̥)
- sā śaktā “She is capable” (śak)
In general, however, the form in Ktáḥ only refers the action to the past in the vaguest terms. When used predicatively, it can often be translated by a range of English past tenses depending on the context, for example:
ōdanaṁ bhuktaṁ dēvadattēna
translating it as an active sentence in English:
- “Dēvadatta ate the rice.”
- “Dēvadatta has eaten the rice.”
- “Dēvadatta had eaten the rice.”
- “The rice was eaten by Dēvadatta.”
- “The rice has been eaten by Dēvadatta.”
- “The rice had been eaten by Dēvadatta.”
See Speijer §361.
Regarding the formation of the verbal adjective in -tá-, we can say that:
- the suffix is added directly to the verbal root;
- some roots take the augment -i between the root and the affix, which in certain cases is lengthened to -ī; see the section on sēṬ and aniṬ roots for more;
- the root stands in its weakest or zero-grade form (see the section on vowel gradation above); that is to say:
- there is no guṇáḥ or vŕ̥ddhiḥ of the root vowel;
- the root undergoes saṁprasā́raṇam if it is subject to it;
- for certain roots ending in -am or -an, the stem to which -tá- is added ends in -a (since a, from an earlier vocalic m̥ or n̥, is the “zero grade” of the sequence am);
- for other roots ending in -am or -an, the stem to which -tá- is added takes the form -ām or -ān (these are roots which historically ended in a nasal followed by a laryngeal);
- an exception is jan “be born,” which takes the form jā- before the suffix;
- if the root is taught as ending in a long vowel, that probably represents a laryngeal-final root, and such roots generally end in -ī or -i in the weakest grade (representing -H).
In the feminine, the affix -tá- simply has a long vowel, thus -tā́- (formed with the feminine suffix ṬāP).
Several examples of Ktáḥ affixed to roots with different phonological shapes follow.
- Ktáḥ after a root ending in a vowel:
- kr̥ → kr̥táḥ “done”
- nī → nītáḥ “led”
- śru → śrutáḥ “heard”
- bhū → bhūtáḥ “become”
- Ktáḥ after a root ending in a consonant:
- tyaj → tyaktáḥ “abandoned”
- muc → muktáḥ “abandoned”
- vr̥dh → vr̥ddháḥ “grown”
- naś → naṣṭáḥ “destroyed”
- Ktáḥ after a root subject to samprasāraṇa, including roots ending in -am:
- prach → pr̥ṣṭáḥ “asked”
- vyadh → viddháḥ “pierced”
- svap → suptáḥ “slept”
- gam → gatáḥ “gone”
- yaj → iṣṭáḥ “sacrificed”
- Ktáḥ after roots that historically ended in a -nH or -mH:
- śam → śāntáḥ “quieted”
- kam → kāntáḥ “beloved”
- dhvan → dhvāntáḥ “sounded”
- Ktáḥ after roots that historically ended in a laryngeal without a preceding nasal:
- gai → gītáḥ “sung”
- pā → pītáḥ “drunk”
- sthā → sthitáḥ “stood”
- dhā → hitáḥ “placed”
- Ktáḥ after sēṬ roots:
- kamp → kampitáḥ “shaken”
- pat → patitáḥ “fallen”
- grah → gr̥hītáḥ “taken”
- prath → prathitáḥ “spread”
Finally, mention must be made of the substitution of -tá- by -ná- after certain roots. Pāṇini teaches this substitution in the following cases (8.2.42–44):
- roots ending in the consonant -d:
- bhid + Ktá [ + am] → bhinnám “[it was] broken”
- chid + Ktá [ + am] → chinnám “[it was] cut off”
- roots that had a semivowel (l, r, or v) followed by a laryngeal in Indo-European, which in synchronic terms includes:
- roots ending in -lā (-lH, where C stands for any consonant), e.g.:
- mlā + Ktá [ + am] → mlānám “[it was] withered”
- glā + Ktá [ + am] → glānám “[it was] tired out”
- roots ending in īr or ūr (Indo-European -rH); note that this will always retroflex the following n), e.g.:
- str̥̄ + Ktá [ + am] → stīrṇám “[it was] strewn”
- pr̥̄ + Ktá [ + am] → pūrṇám “[it was] filled up”
- a laundry-list of other verbs, most of which can be reconstructed as ending in -uH or -iH in the zero-grade in Indo-European:
- lū + Kta [ + am] → lūnám “[it was] cut” (Indo-European luH-nóm)
- dhū + Ktá [ + am] → dhūnám “[it was] shaken” (Indo-European dʰuH-nóm)
- lī + Ktá [ + am] → līnám “[it was] dissolved” (Indo-European liH-nóm)
- vij + Ktá [ + sU] → vignáḥ “[he was] shaken”
- lag + Ktá [ + am] → lagnám “[it was] stuck”
- ruj + Ktá [ + am] → rugṇám “[it was] destroyed”
- roots ending in -lā (-lH, where C stands for any consonant), e.g.:
Another verbal adjective can be formed by adding the suffix that Pāṇini calls KtávatU, i.e., -távat-, which can be considered a combination of -tá- (the past verbal adjective) and -vat- (the possessive suffix). It is declined exactly like stems in -vat, that is, with the stem -vant- in the strong cases and -vat- in the weak cases, and with the stem -vatī- (KtávatU + ṄīP) in the feminine.
Like Ktáḥ, KtávatU refers to the past. Unlike Ktáḥ, it always agrees with the agent kartŕ̥ of the verb, rather than with the patient kárma. Although, as adjectives, they can be used to qualify a noun can be used in any gender, number, and case, they are most commonly used as predicate adjectives. Since intransitive verbs already reflect the agent in their tá-participle, the távat-participle is mostly used to refer to the agent of transitive verbs.
- rāmō rāvaṇaṁ hatavān “Rāma has killed Rāvaṇa.”
- sītā rāmaṁ dr̥ṣṭavatī “Sītā has seen Rāma.”
- pāṇḍavā gatavantaḥ “The Pāṇḍavas have gone.”
- kumbhakāraḥ kumbhaṁ kr̥tavān “The potter has made a pot.”
Pāṇini calls these formations kr̥tyāḥ Aṣṭādhyāyī 3.1.95, after one of the main types discussed below. Like the verbal adjectives in Kta, they generally have an absolutive alignment pattern: they refer to either the agent of an intransitive verb or the patient of a transitive verb. Although often called the “future” verbal adjective, or “future” participle, these suffixes form adjectives that often have a potential force, or an optative, permissive, or imperative force.
As to their formation, all of these suffixes added directly onto the root, with differences in the gradation of the root.
Regarding their syntax, it should be noted that, in contrast to the past absolutive participle Kta, the agent kartr̥ of kr̥tyāḥ is often expressed in the genitive (षष्ठी) in addition to the instrumental (तृतीया):
- vandyaiḥ puṁsāṁ raghupatipadair aṅkitaṁ mēkhalāsu “marked on its slopes with the footprints of Rāma, which are to be worshiped by men.” Mēghadūtam 12
- mayāvaśyaṁ dēśāntaraṁ gantavyaṁ “I absolutely need to go to another country” Pañcatantram 167 (from Speijer)
One important point to be noted in this connection is the kr̥tyaḥ of bhū, and especially the form bhavitávyam. This is very often used in the bhāvḗ prayōgáḥ, or impersonal construction, which means that the agent of the verbal action will be put into the instrumental rather than the nominative:
- asmin latāmaṇḍapē saṁnihitayā tayā bhavitavyam “she has to be close to this bower” Śākuntalam 3 (from Speijer)
See Speijer §357.
Before these suffixes (Aṣṭādhyāyī 3.1.96), the root stands in the full grade. SēṬ roots take the augment iṬ in this form.
The difference between these two suffixes is simply that the accent is on the first syllable of the suffix in the case of tavya, whereas in the case of tavyaT, the anubandha T causes the final syllable to have a svarita accent. Historically tavyaT derived from a suffix -tavía.
- kr̥ + tavyaT [ + am] → kartavyàm “[it is] to be done”
- sthā + tavya [ + am] → stātávyam “[one must] stand”
- ji + tavya [ + am] → jētávyam “[it is] to be conquered”
- bhū + tavya [ + am] → bhavitávyam “[it] must be”
- man + tavya [ + sU] → mantávyaḥ “[it] is to be thought about”
These forms both involve a suffix -ya-. The difference is that yaT induces full grade of the root syllable, whereas ṆyaT induces a gradational pattern that we have been calling Ṇ-vr̥ddhi, where the root takes the lengthened grade or vŕ̥ddhiḥ if it ends in a vowel, and the full grade or guṇáḥ if it ends in a consonant.
The suffix yaT is added:
- after most roots ending in a vowel (Aṣṭādhyāyī 3.1.97) apart from those in r̥, e.g., dā, śru;
- after most roots ending in a labial sound preceded by a, e.g., tap, labh, as well as a few other roots with a CVC pattern, such as śak, sah and (when not preceded by a preverb) gad, car, etc. (this specification is necessary because most roots ending in a consonant take the suffix ṆyaT);
ṆyaT is only taught for roots that end in the vowel r̥ or a consonant.
Note that both yaT and ṆyaT are affixes beginning with the phoneme y, and they therefore fall under the scope of the special sandhi rule Aṣṭādhyāyī 6.4.79: before such a suffix, the guṇáḥ and vŕ̥ddhiḥ vowels are replaced by their corresponding vowel + semivowel segments (ē → ay, ō → av, ai → āy, au → āv).
- dā + yaT [ + am] → dḗyam “[it is] to be given” (for the substitution of the root-final ā with ē, see Aṣṭādhyāyī 6.4.65).
- śru + yaT [ + am] → śrávyam “[it is] to be heard”
- kr̥ + ṆyaT [ + am] → kāryàm “[it is] to be done”
The root is in its “weakest” grade (see vowel gradation above). In addition, if the root by itself does not constitute a heavy syllable (i.e., if it ends in a short vowel), then the augment -t- is inserted between the root and the affix.
- kr̥ + KyaP [ + am] → kr̥tyám “[what is] to be done”
- stu + KyaP [ + sU] → stutyáḥ “[one who is] to be praised”
- han + KyaP [ + sU] → hatyáḥ “[one who is] to be killed”
- śās + KyaP [ + sU] → śiśyáḥ “[one who is] to be taught”
The root takes the full grade before this suffix, which is accented on the penultimate syllable (as indicated by the diacritic R):
- kr̥ + anīyaR [ + am] → karaṇī́yam “[it is] to be done”
- bhū + anīyaR [ + am] → bhavanī́yam “[it is] to be”
- vac + anīyaR [ + am] → vacanī́yam “[it is] to be spoken of”
Note that while anīyaR usually requires guṇa of the root vowel, just as in the case of the guṇa of a root vowel before the present tense forming suffix ŚaP, guṇa does not take place if the root ends in (1) a long vowel followed by a consonant, or (2) two consonants.
The root takes Ṇ-vr̥ddhi (see above). The form of the suffix is -in-. There is a small set of roots to which this suffix can be added, on its own, to form agent nouns:
- grah → grāhī “grasper”
More commonly, this suffix is attached to a verbal root at the end of an upapada compound (see below) and has a habitual meaning (tācchīlyē “in the sense of having that as one’s habitual activity”). For example:
- yaj → darśapūrṇamāsayājī “one who performs the darśapūrṇamāsa sacrifice”
- kruś → uṣṭrakrōśī “given to making noises like a camel”
The root takes “Ṇ-vr̥ddhi” (see the section on vowel gradation above).
- kr̥ → kā́rakaḥ “doer”
- hr̥ → hā́rakaḥ “taker”
Taught in Aṣṭādhyāyī 3.1.134. The suffix -ana-, without any change in the stem. The closely-related suffix LyuṬ (see below) is identical in form, but has the sense of an instrument, or the verbal action, rather than an agent.
- nand → nandanaḥ “that which makes rejoice, a son”
- sādh → sādhanaḥ “that which makes accomplished, a means”
Action nouns are nouns which refer to the verbal action bhāvaḥ rather than to one or another of the participants in the action.
This is a suffix -ti- which is added to a root in the zero grade (as required by the anubandha K). It forms feminine verbal nouns. Because of the anubandha N, these derivatives are accented on the initial syllable. See Aṣṭādhyāyī 3.3.94 and following.
- man- + KtiN ( + sU) → mátiḥ “thought, intelligence”
- āp + KtiN ( + sU) → ā́ptiḥ “attainment”
- sādh + KtiN ( + sU) → síddhiḥ “accomplishment”
- kram + KtiN ( + sU) → krā́ntiḥ “overcoming”
The suffix -ana-, which ias added to a root in the full grade, forms neuter verbal nouns that are accented on the syllable preceding the suffix. The same suffix is also used to form nouns expressive of an instrument karaṇam or location adhikaraṇam of a verbal action.
- gam + LyuṬ ( + sU) → gámanam “going”
- kr̥ + LyuṬ ( + sU) → káraṇam “doing”
- śās + LyuṬ ( + sU) → śā́sanam “teaching”
- paṭh + LyuṬ ( + sU) → páṭhanam “reciting”
- adhii + LyuṬ ( + sU) → adhyáyanam “studying”
The suffix -a-, which is added to a root in the full grade, forms masculine verbal nouns that are accented on the root syllables. The marker GH means that a root-final palatal is converted into the corresponding velar before the suffix Aṣṭādhyāyī 7.3.52. The root goes into the full grade or guṇáḥ, unless it ends in a vowel, or has the vowel a followed by a single consonant, in which case it takes vŕ̥ddhiḥ (see Ṇ-vŕ̥ddhiḥ).
- bhū + GHaÑ ( + sU) → bhāváḥ “going”
- viśiṣ + GHaÑ ( + sU) → viśēṣáḥ “difference”
- paṭh + GHaÑ ( + sU) → pāṭháḥ “reciting”
- yuj + GHaÑ ( + sU) → yōgáḥ “joining”
- pac + GHaÑ ( + sU) → pākáḥ “cooking”
- ruj + GHaÑ ( + sU) → rōgáḥ “sickness”
A “secondary derivational suffix,” or taddhitá suffix, is one that is added onto an existing nominal stem prātipadikam in order to form a new nominal stem. The suffix, and thus the resulting form, generally expresses a relation to the meaning of the stem to which it is added.
We can distinguish the following types of taddhitá suffixes:
Gender in Sanskrit is partly inflectional (expressed through endings) and partly derivational (expressed through differences in the stem to which endings are added). One example of the inflectional expression of gender is the use of special endings when a word is used in the feminine: for instance, the adjective śuciḥ, meaning “pure,” can qualify a noun in the masculine or feminine gender, but the genitive form śucyāḥ is only used to qualify a noun in the feminine gender (whereas śucēḥ can be used to qualify either a masculine or feminine noun).
Gender being expressed derivationally are found when forms based on one stem are used in the masculine/neuter, and forms based on another stem are used in the feminine. To express this paradigmatic relationship, Sanskrit grammarians say that the feminine stem is formed by adding a feminine stem forming suffix strīpratyayaḥ onto the masculine/neuter stem Aṣṭādhyāyī 4.1.3. (We could of course say the reverse, namely, that the masculine/neuter stem is formed by adding a suffix onto the feminine stem, but it is generally more convenient to proceed in the other direction.)
The primary feminine stem forming suffixes are ā and ī, which Pāṇini teaches as the following:
- āP, a general term for the suffix ā, which actually includes three suffixes:
- ṬāP Aṣṭādhyāyī 4.1.4. The general suffix ā, used when the corresponding masculine/neuter stem ends in a.
- ḌāP Aṣṭādhyāyī 4.1.13. The suffix ā, which replaces the final an of a masculine/neuter stem.
- CāP Aṣṭādhyāyī 4.1.75. The suffix ā, but with a final udā́ttaḥ accent. Used in patronymics and metronymics, e.g., kausalyā́.
- ṄīP Aṣṭādhyāyī 4.1.5. The suffix ī, appended to a stem in r̥ or n, as well as most stems ending in t (technically prescribed when the corresponding masculine/neuter stem is formed with a suffix that has the anubandhas u, r̥, or l̥, hence the present participle [ŚatR̥] and the possessive adjectives in mat and vat [matUP]), and to other stems in special circumstances.
- rājan-ṄīP → rājñī “queen”
- ṄīṢ Aṣṭādhyāyī 4.1.25. The suffix ī, with a final udā́ttaḥ accent. It is used with the augment ān ānUK to form indrāṇī́, yavanānī́, etc.
- gaura-ṄīṢ → gaur–ī́ “white”
- sundara-ṄīṢ → sundarī́ “beautiful”
An important set of suffixes form abstract nouns from an existing nominal stem. Such derivatives can be translated, on a first pass, as “x-ness,” “being x,” and so on. But these nouns are very often used as predicates, especially in certain styles of Sanskrit. Hence they can often mean “the fact [that something] is x,” or “[something’s] being x.” In such constructions, these nouns very often take an adnominal modifier—usually expressed in the genitive case—to refer to the thing that possesses “x-ness,” or stated more simply, the thing that is said to be x.
- tva Aṣṭādhyāyī 5.1.119. The suffix -tvá-, without any change in the stem. The resulting noun is neuter.
- nī́la-tvá- → nīlatvám “blueness,” gaganasya nīlatvam “the sky’s blueness, the sky’s being blue.”
- tal Aṣṭādhyāyī 5.1.119. The suffix -tā́-, without any change in the stem. The resulting noun is feminine.
- nī́la-tā́- → nīlatā́ “blueness,” gaganasya nīlatā “the sky’s blueness, the sky’s being blue.”
- ṢyaÑ Aṣṭādhyāyī 5.1.123. The suffix -ya-, which causes vŕ̥ddhiḥ of the initial syllable of the stem. The accent is on the initial syllable (due to Ñ). The resulting noun is neuter. There is a list gaṇaḥ of words to which this suffix can be added:
- śuklya-ṢyaÑ- → śáuklyam “whiteness”
- dr̥ḍha-ṢyaÑ- → dā́rḍhyam “firmness”
- kr̥ṣṇa-ṢyaÑ- → kā́rṣṇyam “blackness”
- aṆ Aṣṭādhyāyī 4.1.83. The suffix -a-, with “Ṇ-v̥rddhi” of the stem (see vowel gradation). The resulting noun is neuter.
- gurú-aṆ- → gaúravam “heaviness, seriousness”
- laghú-aṆ- → lā́ghavam “lightness”
One important class of suffixes indicates that someone possesses that to which the suffix is added. Usually these suffixes refer to inalienable possession, wherein the possessed object comes to characterize the one who possesses it.
The suffix -in- (feminine -inī-). The final vowel of the stem is dropped before this suffix.
- jñāna-inI- → jñānin- “possessed of knowledge, knowledgeable, wise”
- kuṭumba-inI- → kuṭumbin- “one who possesses a family, a householder”
The suffix -mat- and -vat-. These suffixes are in complementary distribution: vatUP is used whenever the base to which the suffix is added ends in the vowel a, and matUP is used elsewhere. (Note that stems in -n take a form that ends in -a before taddhita suffixes.) These suffixes involve vowel gradation, and examples of their declension have been given above.
The feminine to both matUP and vatUP is formed by adding the feminine suffix ṄīP to the weak form of the stem, thus -matī- and -vatī-.
- dhana-vatUP ( + sU) → dhanavān “wealthy”
- śakti-matUP ( + sU) → śaktimān “powerful”
- ātma-vatUP ( + sU) → ātmavān “self-possessed”
- gō-matUP ( + sU) → gōmān “rich in cows”
By Aṣṭādhyāyī 6.3.119, when the suffix matUP is used to form a name, the final vowel of the stem is lengthened:
- padma-vatUP ( + sU) → padmāvatī
- amara-vatUP ( + sU) → amarāvatī