ek%Aco dve prathamasya ONPANINI 61001
Replace the first onevowel part with two.mmmmmmmmm ^ M+ C+ 495

Sometimes, we have to " reduplicate" roots, that is, we replace the first part of the root with two of the same. The first part is everything up to the first vowel. Examples --

The first part of yudh is yu

so "reduplicate yudh" means "replace yudh with yu + yudh"

The first part of nI is nI

so we replace nI with nI + nI

The first part of ad is a

so we replace ad with a + ad

See also the exceptions below, ajAderdvi and nan;d;rAssaMyo.

You said that "sometimes" we reduplicate. When exactly?

When the rules liTidhA sanyaGoH zlau caGi below say so. Namely, before the affixes liT san yaG zlu caG.

Eample. Rule liTidhA teaches "reduplicate what is before liT". This means that when rakS is before Nal (which is a liT affix), we must replace the ra of rakS with rara --

rakS + Nal →liTidhA rarakS + Nal → rarakSa "he protected"

KAZIKA adhikAro 'yam. ekAcaH iti ca, dve iti ca, prathamasya iti ca tritayam adhikRtaM veditavyam. ita uttaraM yad vakSyAmaH prAk saMprasAraNavidhAnAt tatra ekAcaH prathamasya dve bhavataH ity evaM tad veditavyam. vakSyati liTi dhAtoranabhyAsasya iti. tatra dhAtoravayavayasya anabhyAsasya prathamasya ekAco dve bhavataH. jajAgAra. papAca. iyAya. Ara. ekAca iti bahuvrIhinirdezaH. eko 'c yasya so 'yam ekAcityavayavena vigrahaH. tatra samudAyaH samAsArthaH. abhyantarazca samudAye 'vayavo bhavati iti sAckasya eva dvirvacanaM bhavati. evaM ca pacityatra yena eva acA samudAyaH ekAc, tenA eva tadavayavo 'cchabdaH pazabdazca. tatra pRthagavayavaikAc na dvirucyate, kiM tarhi, samudAyaikAjeva. tathA hi sakRcchAstrapravRttyA sAvayavaH samudAyo 'nugRhyate. papAca ityatra prathamatvaM vyapadezivadbhAvAt. iyAya, Ara ityatra ekActvam api vyapadezibhAvAdeva. dviHprayogazca dvirvacanam idam. AvRttisaGkhyA hi dve iti vidhIyate. tena sa eva zabdo dvirucAryate, na ca zabdAntaraM tasya sthAne vidhIyate.

aj-Ader dvitIyasya ONPANINI 61002
(But double) the second (onevowel part) of a vowelstarter.mmmmmmmmm ^ C+ 496

Exception to ekAcodve. The roots that start with a vowel and have at least two vowels replace their second one-vowel part with two.

Example. The root aTiSa means "be about to wander" and starts with a vowel. When it reduplicates, we change --

a Ti Sa a Ti Ti Sa → aTiTiSa "want to wander"

as in --

aTiSa + laT tip aTiTiSati "he wants to go walkabout"

See exception nandrA right below.

KAZIKA prathamadvirvacanApavAdo 'yam. ajAder dvitiyasya ekaco dvirvacanam adhikriyate. acAdir yasya dhAtoH tadavayavasya dvitIyasya ekAco dve bhavataH. aTiTiSati. aziziSati. aririSati. arteH smipUGraJjvazAM sani 72074 iti iT kriyate. tasmin kRte guNe ca raparatve ca dvirvacane 'ci 12059 iti sthAnivadbhAvaH prApnoti. tatra pratividhAnaM dvirvacananimitte 'ci iti ucyate. na ca atra dvirvacananimittamiT. kiM tarhi? kAryI. na ca kAryI nimittatvena azrIyate. tathA hi kGinimittayor guNavRddhyoH pratiSedho vidhIyamAnaH zayitA ityatra na bhavati, na hi kAryiNaH zIGo guNaM prati nimittabhAvaH iti. atra kecidajAdeH iti karmadhArayAt paJcamIm icchanti. ac ca asau Adizca ityajAdiH, tasmAtajAderuttarasya ekAco dve bhavataH iti. teSAM dvitIyasya iti vispaSTarthaM draSTAvyam.

That root aTiSa does not appear in my list of roots, where did you get it from?

I added san after root aT "wander". This makes the root aTiSa --

aT + san → aT + iT + san →kric aTiSa

That's complicated. Shouldn't you have used a root from the dhAtupATha for your example?

dhAtupATha roots that start with a vowel and have another vowel are very uncommon.

Well, UrNu is such a root. You could use that one as an example.

No, I cannot. UrNu is covered by the exception nandrA.

na n;d;rAH saMyogAdayaH ONPANINI 61003
But not the n d r b that start a cluster.mmmmmmmmm ^ C+ 497

Exception to ajAderdvi. If the second part of a vowel-starter root starts with n d r b plus other consonants, we don't reduplicate that n d r b, but just the rest of the second part.

Examples. The roots undiSa and Urnu have second parts ndi rnu. According to the previous rule we would have replaced ndi with two ndi and rnu with two rnu. But this exception says that we replace one di with two di and one nu with two nu --

u ndi Sa u n di di Sa → undidiSa

U rnu U r nu nu → Urnunu →Natvam UrNunu

As in --

und + san + laT tip →sanyaGoH undidiSa + tip undidiSati "he wants to wetten"

Urnu + Nal →liTidhA Urnunu + Nal →acoJNiti Urnunau + a →ecoy UrnunAva →Natvam UrNunAva "he hid"

KAZIKA (laghu) acaH parAH saMyogAdayo nadarA dvir na bhavanti. nuzabdasya dvitvam. UrNunAva. UrNunuvatuH. UrNunuvuH.

Why do we say "n d r b only"?

Root IkSiSa has a second part kSi, and all of it reduplicates, including the k --

IkS + san → IkS + iSan → IkSISa →ajAderdvi IkSikSiSa →halAdi IkikSiSa →kuhozcuH IcikSiSa "want to see"

What's that Urnu thing? I though that the root was UrNu "conceal".

The root is listed in the dhAtupATha as Urnu, but everybody calls it UrNu anyway, because Urnu is a tonguetwister.

Where's the b in your translation coming from? The original says nandrAH, not nandbrAH.

It comes from a vArttika. This b turns ubj + san into ubjijiSa "want to opress".

pUrvo 'bhyAsaH ONPANINI 61004
stammer is the first (part).mmmmmmmmm ^ M+ C+ 498

When we replace one thing with two by the above rules, the first copy is called the stammer.

Example:

the stammer of papac is pa

ubhe: abhyastam ONPANINI 61005
stammered are both.mmmmmmmmm ^ M- C+ 499

When we replace one part of a root with two by the above rules, the new root is a stammered.

Example. When ekAcodve replaces root pac with papac, by making one pa into two pa,

the stammer is the first pa

and the stammered is papac

There are two sorts of stammered --

(1) Those made by this rule (such as papac).

(2) Those mentioned in the next rule, jakSityA (such as jAgR).

KAZIKA dve iti vartamAne ubhegrahaNaM samudAyasaJjApratipattyartham. ye eve vihite te ubhe api samudite abhyastasaMjJe bhavataH. dadati. dadat. dadhatu. ubhegrahaNaM kim? nenijati ityatra abhyastAnAm AdiH iti samudAye udAttatvaM yathA syAt, pratyekaM paryAyeNa vA mA bhUtiti. abhyastapradezAH abhyastAnAm AdiH 6-1-189 ityevam AdayaH.

jakSityAdayaH SaT ONPANINI 61006
jakSa~ jAgR daridrA cakAsR zAsu dIdhIG vevIG (are stammered).mmmmmmmmm ^ C+ 500

These seven special roots are stammered just because this rule says so, even though they weren't made by reduplicating other roots. They are all lukclass.

Being a stammered makes rule adabhyastAt work. Example --

jAgR + laT jhi jAgR ( stammered ) + ati''' →ikoyaNaci jAgrati "they are awake"

jAgR + laT tip →jAgrovi jAgarti "is awake"

It also prohibits the num from nAbhyastAccha. So daridratam "(him) that is poor", like dadatam.

Why is zAs in this list? Unlike the others, it doesn't even remotely look like a reduplicate.

So that jhontaH won't work here --

zAs + laT jhi →adabhyastAt zAs + ati → zAsati "they command"

The original rule means "jakS and the next five", but you say this affects jakS and the next six, namely the seven roots starting at ad 02.0066. Why didn't pANini say jakSityAdayassapta?

He dozed off, like Homer. Rest assured that he meant to include vevIG. All grammarians agree on that.

Didn't you say last month in class that he never ever made a mistake?

See. In order to show respect to teachers, we never ask impertinent questions like that one in public. Welcome to Indian tradition.

KAZIKA abhyastam iti vartate. jakSa ity ayaM dhAtuH ityAdayazca anye SaT dhAtavaH abhyasta-saMjJA bhavanti. seyaM saptAnAM dhAtUnAm abhyasta-saMjJA vidhIyate. jakSa bhakSahasanyoH ityataH prabhRti vevIG vetinA tulye iti yAvat. jakSati. jAgrati. daridrati. cakAsati. zAsati. dIdhyate, vevyate ityatra abhyastAnAm AdiH ityeSa svaraH prayojanam. dIdhyat iti ca zatari vyatyayena sampAdite n' .AbhyastAc chatuH iti numaH pratiSedhaH.

liTi dhAtor an-abhyAsasya ONPANINI 61008
Before liT, ( reduplicate) stammer-less root.mmmmmmmmm ^ M+ C+ 501

Examples --

pac + liT tip → pac + Nal →ata::upa pAc + Nal pApAca →hrasvaH papAca "he cooked"

kSip + Nal →puganta kSep + Nal kSekSepa cikSepa "threw"

jAgR + Nal →acoJNiti jAgAr + Nal jAjAgAr + a →hrasvaH jajAgAra "woke up"

vac + Nal →ata::upa vAc + a vA + vAc + a →liTyabhyAsa uvAca "said"

What about pAci, from pac + Nic ?

Cannot reduplicate the sanAdyanta. See kAs;pratyayAdAmamantreliTi.

san;yaGoH ONPANINI 61009
( reduplicate) before san and yaG.mmmmmmmmm ^ M- C+ 502

Examples --

pac + san + laT tip papakSa + ti →sanyataH pipakSa + ti → pipakSa + zap + ti →atoguNe pipakSati "wants to cook"

pac + yaG + laT ta pa + pac + yaG + ta → papacyate →dIrghokitaH pApacyate "cooks a lot, is always cooking"

yudh + san yuyutsa

Why is there no puganta in yuyutsa?

Because after ik-plus-hal, iT-less san is kit. .

zlau ONPANINI 61010
( reduplicate) before zlummmmmmmmm ^ M+ C+ 503

The zluclass roots get zlu added. Then the zlu reduplicates the root by this rule, and disappears by pratyayasyaluk;zlu;lupaH.

hu + laT tip → hu + zlu + tip hu + hu + zlu + tip →pratyayasyaluk;zlu;lupaH hu hu tip →hardsoft hu + ho + ti →kuhozcuH juhoti "sacrifices"

bhI + laT jhi bibhI + jhi → bibhI + ati''' →ikoyaNaci bibhyati "they fear"

bhI + laT tip →hardsoft bhe + ti bebhe + ti →hrasvaH bibheti "fears"

dA + laT tip dA + dA + tip →hrasvaH dadAti "gives"

dA + laT jhi dadA + jhi → dadA + ati''' →znAbhya dadati "they give"

caGi ONPANINI 61011
( reduplicate) before caGmmmmmmmmm ^ C+ 504 luG

The caG affix appears in some luG. Examples --

dru + luG tip →Nizridrusru dru + caG + tip →luGlaG aT + dru + caG + tip → a + dru + caG + t''' aT + dudru + caG + t''' → adudruvat "ran"

pAci + luG tip apIpacat "he made (them) cook"

kAmi + luG tip acIkamat

narti + luG tip anInRtat

cori + luG tip acUcurat

nAyi + luG tip anInayat

The caGi reduplication has some peculiarities --

every vowel except u U reduplicates as i, by sanvalla

The Nic affix disappears by NeraniTi,

the root vowel shortens by NaucaGyu,

and the vowel of the stammer is often long -- I or U by sanvalla sanyataH dIrgholaghoH.

KAZIKA caGi parato 'nabhyAsasya dhAtoravayavasya prathamasya ekAco dvitIyasya vA yathAyogam dve bhavataH. apIpacat. apIpaThat. ATiTat. Azizat. Ardidat. pacAdInAM NyanatAnAm caGi kRte NilopaH, upadhAhrasvatvaM, dvirvacanam ityeSAM kAryANAM pravRttikramaH. tathA ca sanvallaghuni caGpare iti sanvadbhAvo vidhIyamAno hrasvasya sthAnivadbhAvAn na pratiSidhyate. yo hyanAdiSTAdacaH pUrvaH tasya vidhiM prati sthAnivadbhAvo bhavati. na ca asmin kAryANAM krameNAniSTAdacaH pUrvo 'bhyAso bhavati iti. ATiTatiti dvirvacane 'ci 11059 iti sthAnivadbhAvAt dvitIyasya ekAcaH dvirvacanaM bhavati.

vaci;svapi;yajAdInAM kiti ONPANINI 61015
( stretch) vac, svap, yaj-class before kit.mmmmmmmmm ^ M+ C+ 505 stretch

The yaj-class roots are the last nine of the zapclass ,from 01.1157 yaj onwards --

yaj vap vah

vas' veJ vyeJ

hveJ vad zvi "swell"

Examples of vac svap yajjclass stretching before kit --

vac + kta ucta →coHkuH ukta- "said"

svap + kta supta- "asleep, slept"

yaj + kta ij- + ta →vrazca;bhrasja iS + ta →STunA iSTa- "was sacrificed"

yaj + liT mas → yaj + ma →kRsR yaj + ima →asaMyogAlliTkit yaj + ima(k) ij + ima →liTidhA i + ij + ima →akassa Ijima "we sacrificed"

Examples of not stretching before Git --

vac + anti''' → vacanti "they say"

svap + anti''' → svapanti "they sleep"

yaj + caG + mas → ayIyajAma "we sacrificed"

KAZIKA samprasAraNam iti vartate. SyaGaH iti nivRttam. vaci vaca paribhASaNe, bruvo vacir iti ca. svapi JiSvapa zaye. yajAdayaH yaja deva-pUjA-saMgati-karanadAnesu ityataH prabhRti AgaNAntAH. teSAM vacisvapi yajAdInAM kiti pratyaye parataH samprasAraNam bhavati. vaci uktaH. uktavA. svapi suptaH. suptavAn. yaja iSTaH. iSTavAn. vapa uptaH. uptavAn. vaha UDhaH. UDhavAn. vasa uSitaH. uSitavAn. veJutaH. utavAn. vyeJ saMvItaH ( halaH). saMvItavAn. hveJ AhUtavAn. vada uditaH. uditavAn. Tuozvi zUnaH. zUnavAn. dhAtoH svarUpa-grahaNe tatpratyayekAryaM vijJAyate. teneha na bhavati, vAcyate, vAcikaH iti.

grahi;jyA;vayi; vyadhi;vaSTi;vicati; vRzcati;pRcchati;bhRjjatInAM Giti ca ONPANINI 61016
( stretch) grah jyA vay, vyadh vaz vyac, vrazc pracch bhrasj before kGit.mmmmmmmmm ^ M- C+ 506 stretch

The nine grahijyA roots are --

grah "grab, take" jyA "grow old" 01.0034 vay "go" 09.0547

vyadh "pierce" vaz "command, coerce" vyac

vrazc pracch "ask" bhrasj "fry"

They stretch before kit and Git affixes.

Examples of stretching before the kit affixes ktvA and kta --

grah + ktvA gRh + tvA →ArdhadhAtukasyeDvalAdeH gRh + itvA →graholiTi gRhItvA "after taking"

jyA + kta ji + kta →lvAdi ji + na →halaH jIna- "grown old"

Examples of stretching before the Git affixes znA jhi za --

grah + laT tip → grah + znA + tip gRhnAti →Natvam gRhNAti "grabs"

jyA + znA + tip jinAti "grows old"

vaz + jhi → vaz + anti''' uzanti "they command"

pracch + za + tip pRccha + tip → pRcchati "asks"

In those examples, znA, jhi and za are Git because of rule hard apit is Git .

KAZIKA graha upAdAne, jyA vayohAnau, veJo vayiH 24041, vyadha tADane, vaza kAntau, vyac vyAjIkaraNe, ovrazcU chedane, praccha jJIpsAyAm, bhrasja pAke ityeteSAM dhAtUnAM GIti pratyaye parataH cakArAt kati ca saMprasAraNaM bhavati. graha gRhItaH. gRhItavAn. Giti gRhNAti jarIgRhyate. jyA jInaH. jInavAn. lvAdibhya iti niSThAnatvam. Giti jinAti. jejIyate. halaH iti samprasAraNadIrghe kRte pvAdInAM iti hrasvaH kriyate. vayi liTi parataH veJo vayiH AdezaH, tasya GidabhAvAt kidevodAhriyate. UyatuH. UyuH. yadyevaM vayigrahaNam anarthakam, yajAdiSu veJ paThyate? na evaM zakyam. liTi tasya veJaH 61040 iti pratiSedho vakSyate, tatra yathA eva sthAnivadbhAvAd vyervidhiH evaM pratiSedho 'pi prApnoti? na eSa doSaH. liTi vayo yaH 61038 iti yakArasya saMprasAraNapratiSedhAd vayervidhau grahaNaM pratiSedhe cAgrahaNamanumAsyate? satyam etat. eSa eva arthaH sAkSAn nirdezena vyeH spaSTIkriyate. vyadha viddhaH. viddhavAn. Giti vidhyati. vevidhyate. vaza uzitaH. uzitavAn. Giti uSTaH. uzanti. vyaca vicitaH. vicitavAn. Giti vicati. vevicyate. vyaceH kuTAditvamanasi pratipAditam, tena sarvatra aJNiti pratyaye samprasAraNaM bhavati, udvicitA, udvicitum, udvicitavyam iti. vrazceH vRkNaH. vRkNavAn. atha katham atra kutvaM, vrazcabhrasja iti hi Satvena bhavitavyam? niSThAdezaH SatvasvarapratyayavidhIDvidhiSu siddho vaktavyaH. tatra StvaM prati natvasya siddhatvAd jJalAdir niSThA na bhavati. kutve tu kartavye tadasiddham eva iti pravartate kutvam. Giti vRzcati. varIvRzcyate. praccha pRSTaH. pRSTavan. Giti pRcchati. parIpRchyate. praznaH. naGi tu prazne c' Asanna-kAle iti nipAtanAdasaMprasAraNam. bhrasja bhRSTaH. bhRSTavAn. Giti bhRjjati. barIbhRjyate. sakArasya jhalAM jaz jhazIti jaztvena dakAraH, stoH zcunA zcur iti zcutvena jakAraH.

liTy abhyAsasy' .obhayeSAm ONPANINI 61017
Before liT, ( stretch) stammer of the vacisvapi and the grahijyA.mmmmmmmmm ^ C+ 507 stretch

This rule only works when the root has not been stretched, that is to say, when the liT is akit, as in --

vac + Nal →acoJNiti vAc + Nal →liTidhA vA + vAc + a u + vAc + a → uvAca "he said"

yaj + Nal yA + yAj + Nal i + yAj + Nal → iyAja "he sacrificed"

See the other examples in the kAzikA.

This rule works before halAdizzeSaH can work --

vyac + Nal →ata_up vyAc + a vya + vyac + Nal vivyAca "it extended"

vyadh + Nal vivyAdha "he pierced"

KAZIKA ubhayeSAM vacyAdInAM grahyAdInAM ca liTi parato 'bhyAsasya samprasAraNaM bhavati. vaci uvAca. uvacitha. svapi suSvApa. suSvapitha. yaja iyAja. iyajitha. Duvap uvApa. uvapitha. grahyAdInAm tatra graher avizeSaH. jagrAha. jagrahitha. jyA jijyau. jijyitha. vayi uvAya. uvayitha. vyadha vivyAdha. vivyadhitha. vaza uvAza. uvazitha. vyaca vivyAca. vivyacitha. vRzcateH satyasati vA yoge na asti vizeSaH. yogArambhe tu sati yadi samprasAraNam akRtvA halAdizeSeNa repho nivartyate, tadA vakArasya samprasAraNaM prApnoti. atha rephasya samprasAraNaM kRtvA uradatvaM raparatvaM ca kriyate, tadAnI muradatvasya sthAnivadbhAvAt na samprasAraNe samprasAraNam iti pratiSedho bhavati ity asti vizeSaH. vavrazca. vavrazcitha. pRcchati bhRjjatyoravizeSaH. akidarthaM ca idam abhyAsasya samprasAraNaM vidhIyate. kiti hi paratvAd dhAtoH samprasAraNe kRte punaH prasaGgavijJAnAd dvirvacanam, UcatuH, UcuH iti. adhikArAdeva ubhayeSAM grahaNe siddhe punarubhayeSAm iti vacanaM halAdiH zeSam api bAdhitvA samprasAraNam eva yathA syAd iti. vivyAdha.

You say that this rule only works when the root has not been stretched. Why not? What happens if the root has been stretched?

In that case this rule does nothing, because it cannot. The stammer has already been stretched by other rules. See --

vac + us →vacisvapi uc + us →liTidhA u + uc + us →akassa Ucus "they said"

Why does inria conjugation give vavyaca?

Either inria is wrong or the kAzikA is. Ask hyderabad for its vote.

svapi;syami;vyeJAM yaGi ONPANINI 61019
( stretch) svap syam vyeJ before yaG.mmmmmmmmm ^ C+ 508

Only examples --

svap + yaG sup + ya →sanyaGoH susupya →guNoyaG sosupya →kric soSupya "sleep like a log"

syam + yaG sim + ya sesimya "shout like crazy"

vyeJ + yaG vi + ya →akRtsArva vI + ya →sanyaGoH vIvIya →guNoyaG vevIya "do lots of hiding"

KAZIKA JiSvapa zaye, syamu, svana, dhvana zabde, vyeJ saMvaraNe ityeteSAM dhAtUnAM yaGi parataH samprasAraNaM bhavati. soSupyate. sesimyate. vevIyate. yaGi iti kim? svapnak.

na samprasAraNe samprasAraNam ONPANINI 61037
No stretching before stretched.mmmmmmmmm ^ M- C+ 509 stretch

talkaround -- If some rule asks us to stretch, say, yuvan or vya, that have several yaN, only the last yaN will be stretched.

Example. When rule zva;yuva tells us to stretch yuvan-, the va is stretched into u, but the yu stays --

yuvan + TA yu + un + A →akassa yUn + A → yUnA "with the young man"

liTi vayo yaH ONPANINI 61038
The y of vai (does not stretch).mmmmmmmmm ^ C+ 510

vai is the optional substitute of veJ made by veJo vayiH before liT.

This rule allows veJovayiH to make the forms uvAya, Uyatus, Uyus "wove".

KAZIKA na samprasAraNam iti anuvartate. liTi parato vayo yakArasya samprasAraNaM na bhavati. uvAya, UyatuH, UyuH. liD-grahaNam uttar%Artham.

vaz c' .Asy' .AnyatarasyAM kiti ONPANINI 61039
This (y) to v before kit ( liT) optionally.mmmmmmmmm ^ C+ 511

So instead of the Uyatus Uyus made by the previous rule, we may also say Uvatus Uvus.

This rule won't work before akit liT, so we say uvAya "he wove", uvAyitha "thou wovedest" compulsorily.

KAZIKA asya vayo yakArasya kiti liTi parato vakArAdezo bhavati anyatarasyAm. UvatuH, UvuH. UyatuH, UyuH. kiti iti kim? uvAya. uvayitha.

veJaH ONPANINI 61040
veJ (does not stretch before liT)mmmmmmmmm ^ C+ 512

veJ before liT turns into vA by Adeca_u. This exception prevents liTyabhyAsa and vacisvapi, and therefore this vA reduplicates as if it were dA making dadau dadatus dadus --

veJ + Nal →Adeca_u vA + Nal vavau "he wove"

vavatus

vavus

See dadau, dadatus, dadus in the hedgehog for details of the rules that worked here.

KAZIKA liTi ityanuvartate. veJ tantusantAne ityasya dhAtoH liti parataH samprasAraNaM na bhavati. vavau, vavatuH, vavuH. kiti yajAditvAt dhAtoH prAptam akityapi liTyabhyAsasya ubhayeSAm 61017 ityabhyAsasya, ataH ubhayaM pratiSidhyate.

lyapi ca ONPANINI 61041
( veJ gets no stretching) before lyapmmmmmmmmm ^ C+ 513

Exception to vacisvapi. lyap is kit because it replaced ktvA.

Example --

pra + veJ + lyap →Adeca_u pravAya "after starting to weave"

KAZIKA veJaH ity anuvartate. lyapi ca parto veJaH samprasAraNaM na bhavati. pravAya. upavAya. pRthagyogakaraNam uttar%Artham.

jyaz ca ONPANINI 61042
jyA (won't stretch before lyap).mmmmmmmmm ^ C+ 514

jyA is the root 09.0034 jyA jInAti "grow old", that would be ordinarily stretched by grahijyA, because lyap is kit.

Example --

pra + jyA + lyap →Adeca_u prajyAya "after starting to grow old"

KAZIKA lyapi ityeva. jyA vayohAnau ityasya dhAtoH lyapi parataH samprasAraNaM na bhavati. prajyAya. upajyAya.

vyaz ca ONPANINI 61043
vyA (is not stretched before lyap)mmmmmmmmm ^ C+ 515

Exception to vacisvapi, that would have stretched vyeJ "hide" because lyap is kit.

Example --

pra + vye + lyap →Adeca_u pra + vyA + lyap pravyAya "after starting to hide"

KAZIKA lyapi ity eva. vyeJ saMvaraNe ity etasya dhAtoH lyapi parataH samprasAraNaM na bhavati. pravyAya. upavyAya. yoga-vibhAgaH uttar%ArthaH.

vibhASA pareH ONPANINI 61044
(vyA stretches) optionally after pari (before lyap).mmmmmmmmm ^ C+ 516

So ordinarily vacisvapi, because of previous rule, will not stretch the vyA that comes from vyeJ --

pari + vyA + lyap →Adeca_u parivyAya "after wrapping"

But if we take this option --

pari + vyA + lyap →vacisvapi pari + vi + ya →halaH parivIya "after wrapping"

KAZIKA lyapi ca vyaz ca iti anuvartate. parer uttarasya vyeJ ity etasya dhAtoH lyapi parataH vibhASA samprasAraNaM na bhavati. parivIya yUpam, parivyAya. samprasAraNe kRte parapUrvatve ca hrasvasya iti tuka prApnoti, sa halaH 64002 iti dIrghatvena paratvAd bAdhyate.

Why long I in parivIya ?

Rootfinal stretched vowels lengthen by halaH.

Shouldn't hrasvasya piti kRti tuk haveworked first?

Obviously no.

Ad eca upadeze '-ziti ONPANINI 61045
dhAtupATha ec to A unless before zit.mmmmmmmmm ^ C+ 517

The roots that end in an ec in the dhAtupATha, like gai and so, replace it with A before all affixes except the zit affixes.

Examples --

gai + tumun gAtum "to sing"

gai + kta gA + kta →ghumAsthAgA gIta- "sung, song"

Counterexample before a zit --

gai + zap + tip →ecoya gAyati "he sings"

KAZIKA dhAtoH iti vartate. ejanto yo dhAturupadeze tasya akArAdezo bhavati, ziti tu pratyaye na bhavati. glai glAtA. glAtum. glAtavyam. zo nizAtA. nizAtum. nizAtavyam. ecaH iti kim? kartA. hartA. upadeze iti kim? cetA. stotA. aziti iti kim? glAyati. mlAyati. kathaM jagle, mamle? na evaM vijJAyate, zakAra id yasya so 'yaM zititi, kiM tarhi, za eva it zit. tatra yasmin vidhistadAdAvalgrahaNe iti zidAdau pratyaye pratiSedhaH. ez zakArAnto bhavati. aziti iti prasajyapratiSedho 'yam, tena etadAttvam anaimittikaM prAgeva pratyayotpatter bhavati iti, suglaH, sumlaH iti Atazcopasarge 31136 iti kapratyayaH, suglAnaH, sumlAnaH iti Ato yuc ity evam Adi siddham bhavatIti. AkArAdhikArastvayaM nityaM samyateH 61057 iti yAvat.

Will this rulework before ez?.

Ooopsie sorry. I mistranslated. The rule as pANini worded it should work on all affixes that have label z anywhere, but actually it only works on those that have z at the start, such as zap.

sRji;dRzor jhaly am akiti ONPANINI 61058
sRj dRz get a(m) before akit seriousmmmmmmmmm ^ C+ 518

The label m in a(m) makes it into a mitaugment. Therefore, saying that sRj dRz get a(m) is the same thing as saying that we add a after their last vowel. So their R turns into Ra, and then, by ikoyaNaci, into ra. Examples --

dRz + tumun dRaz + tum →ikoyaNaci draz + tum →vrazcabhrasja draS + tum →STunA draSTum "to see"

dRz + tRc dRaz + tR draSTR- "seer"

dRz + luG tip →luGlaG adRz + tip → adRz + sic + tip adRaz + s + tip →ikoyaNaci adraz + s + tip → adraz + s + t''' →sicivRddhiH adrAz + s + t → adrAz + s + It''' →SaDhoHkassi adrAkSIt "saw"

Before kit, we get neither this rule nor puganta --

dRz- + kta → dRSTa-

and before akit non- serious, we get puganta as usual --

dRz + lyuT →yuvoranAkau dRz + lanaT →puganta darzana- "seeing"

You said somewhere else that kvip enders, such as the dRg made from dRz + kvip, may be considered to be roots. That being so, why didn't this rule work in devadRgbhyAm "with two god-seers"?

Because this rule mentions sRj and dRz explicitly, and rules that mention particular roots work only before verb affixes, not before noun affixes .

KAZIKA sRja visarge, dRzir prekSaNe ityetayoH dhAtvoH jJalAdAvakiti pratyaye parataH am Agamo bhavati. sraSTA. sraSTum. sraSTavyam. draSTA. draSTum. draSTavyam. laghUpadhaguNApavAdo 'yam amAgamaH. asrAkSIt. adrAkSIt. sici vRddhiH ami kRte bhavati, pUrvaM tu bAdhyate. jJali iti kim? sarjanam. darzanam. akiti iti kim? sRSTaH. dRSTaH. dhAtoH svarUpagrahaNe tatprataye kAryavijJAnAdiha na bhavati, rajjusRDbhyAm, devadRgbhyAm iti.

dhAtvAdeH SaH saH ONPANINI 61064
Rootstarter S to s.mmmmmmmmm ^ C+ 519

In other words: when the dhAtupATha shows a root that starts with S, the real root starts with s. Such a root is said to be a sa-root. The dhAtupATha lists these roots S just to let us know that rule upasargAt works on them.

Examples. The roots sah "endure" and sic' "sprinkle" are Sa-roots, because the dhAtupATha lists them as Sah 01.0988 and Sic 06.0170. But they actually start with s, not S --

Sah + laT ta sah + ta sahate "endures"

Sic + laT tip sic' + tip sic + za + ti →zemucAdInAm siJcati "sprinkles"

abhi + Sic + laT tip abhi + sic' + tip abhi + sic + za + ti →zemucAdInAm abhisiJcati →upasargAt abhiSiJcati "consecrates as king"

KAZIKA dhAtorAdeH SakArasya sthAne sakArAdezo bhavati. Saha sahate. Sica siJcati. dhAtugrahaNaM kim? soDaza. SoDan. SaNDaH. SaDikaH. AdeH iti kim? kaSati. laSati. kRSati. Adeza pratyayayor ity atra SatvavyavasthArtham SAdayo dhAtavaH kecidupadiSTAH. ke punas te? ye tathA paThyante. athavA lakSaNaM kriyate, ajdantyaparAH sAdayaH SopadezAH smisvidisvadisvaJjisvapitayazca, sRpisRjistRstyAsekSRvarjam. subdhAtuSThivuSvaSkatInAM satvapratiSedho vaktavyaH. SoDIyati. SaNDIyati. SThIvati. SvaSkate. SThivu ityasya dvitIyasthakAraSThakArazca iSyate. tena teSThIvyate, TeSThIvyate iti ca abhyAsarUpaM dvidhA bhavati.

No naH ONPANINI 61065
(Rootstarter) N to n.mmmmmmmmm ^ C+ 520

The Nopadeza roots ("the ones taught with Na") are those that start with N in the dhAtupATha, such as nI "lead" (this nI appears listed as NI(G) in the dhAtupATha).

This rule teaches that we must always replace the initial N of a root with n. Example --

NI(G) + kta nI + ta → nIta- "was led"

In other words: this rule says that no roots start with N, an that those that appear to start with N in the dhAtupATha, actually start with n.

KAZIKA dhAtor AdeH ity anuvartate. dhAtor Ader NakArasya nakAra Adezo bhavati. NIJ nayati. Nama namati. Naha nahyati. dhAtvAdeH ity eva, aNati. sub-dhAtor ayam api neSyate, NakAram icchati NakArIyati. upasargAdasamAse 'pi Nopadezasya ity atra Natva-vidher vyavasthArthaM NAdayo dhAtavaH kecid upadiSyante. ke punas te? ye tathA paThyante. athavA lakSaNaM kriyate, sarve NAdayo NopadezAH, nRtI-nandi-nardi-nakka-nATi-nAthR-nAdhR-varjam.

If NI is the same as nI, why is NI shown with N in the dhAtupATha?

Because rule upasargAdasamAse 'pi Nopadezasya only works on the Nopadeza roots.

My version of the dhAtupATha shows that root as nI, not NIG. How am I supposed to know if a root has N or n in the original dhAtupATha form?

All of the n-starter roots are Nopadeza except nRtI~ nandi nardi nakka nATi nAthR nAdhR.

lopo v;yor vali ONPANINI 61066
v y to lopa before val.mmmmmmmmm ^ C+ 521

Examples --

knUy + kta knUta- "already wet"

knUy + Nic + laT tip → knUy + puk + Nic + tip knU + p + Nic + tip knopayati "wettens"

But that y stays before val letters, such as a of zap or y of yak --

knUy + zap + te''' → knUyate "it is wet"

knUy + yak + te''' → knUyyate "moisture happens"

KAZIKA dhAtor iti prakRtaM yat tad dhAtvAdeH iti punar dhAtugrahaNAn nivRttam. tena dhAtoradhAtozca vakArayakArayoH vali parato lopo bhavati. div didivAn,didivAMsau,didivAMsaH. UyI - Utam. knUyI - knUtam. godhAyA Dhrak 41129. gaudheraH. paceran. yajeran. vakArasya - jIve radAnuk (da. u. 1.163) - jIradAnuH. sriveH - AsremANam. uNAdayo bahulam (33001) iti bahulavacanAc cchvoH zUDanunAsike ca iti UTh na bhavati. vali iti kim? Uyyate. knUyyate. pUrvaM lopagrahaNaM kim? verapRktalopAt pUrvaM vali lopo yathA syAt. kaNDUyateH kvip - kaNDUH. lolUyateH - lolUH. vrazcAdInAm upadezasAmarthyAt vali lopo na bhavati. vRzcati. vavrazca ity atra api hi samprasAraNa-halAdiHzeSayor bahiraGgatvAt prApnoti.

ver apRktasya ONPANINI 61067
One-letter vi affixes lose v.mmmmmmmmm ^ M- C+ 522

These are some of the affected vi affixes --

kvip kvin Nvi cvi

these all lose v by this rule (and lose their other letters because they are all labels).

Example --

indrasya + ji + kvip →supodhA indra + ji + (k)v(ip) →hrasvasyapitikRtituk indra + jit + v indrajit- "defeater of indra"

KAZIKA lopaH iti vartate. ver iti kvib-Adayo vizeSAnanubandhAnutsRjya sAmAnyena gRhyante. ver apRktasya lopo bhavati. brahmabhrUNavRtreSu kvip 62087. brahma-hA. mrUNahA. spRzo 'nudake kvin 62058. ghRtaspRk. tailaspRk. bhajo NviH. ardhabhAk. pAdabhAk. turIyabhAk. apRktasya iti kim? vRdRbhyAM kvin darviH. kR;gR;zR;sR;jAgRbhyaH kvi. jAgRviH.

Why do we bother to say "one-letter only"?

So that this rule won't affect the uNAdi affixes (k)vi(n) and (k)vi, which have two real letters.

(k)vi(n) is seen after vR dR --

dR + kvin → darvi- "wooden, a ladle"

(k)vi appears after kR gR zR sR jAgR --

jAgR + kvi → jAgRvi- "watchful, awake, kindled"

Wait. You mentioned kvin earlier and you said that it always loses v.

Not true. Earlier I said kvin, meaning the affix (k)v(in), which always loses v. The uNAdi affix is (k)vi(n), another affix, and never loses v.

hal;Gy;Abbhyo dIrghAt su;ti;sy;apRktaM hal ONPANINI 61068
su s''' t''' (to lopa) after hal or long GI Ap.mmmmmmmmm ^ M- C+ 523

Examples of su to lopa after hal --

suhRd- + su suhRd "friend" →beforepause suhRt

rAjan- + su →sarvanAmasthAnecA rAjAns rAjAn →nalopaHprA rAjA "king"

Examples of su to lopa after GI Ap --

kukkuTI- + su kukkuTI "chicken"

sItA- + su sItA "pn"

Examples of t''' s''' to lopa after hal ---

dviS + laG tip → aT + dveS + t''' → adveSt adveS →jhalAJjazonte adveD →beforepause adveT "he hated"

dviS + laG sip → aT + dveS + s''' → adveSs adveS →jhalAJjazonte adveD →beforepause adveT "you hated"

Why does the rule say " long" GI Ap? Aren't those two always long?

No. Sometimes gostriyoru shortens them. Then this rule won't work, and the s will stay.

Isn't the hal at the start of this rule useless? If it did not exist, saMyogAnta would erase the su anyway.

You are partially right -- if that hal did not exist, indeed saMyogAnta would make suhRd and rAjAn anyway. HOWEVER. That rAjAn made by saMyogAnta would be unable to trigger nalopaHprA. So we wouldn't get rAjA. Which sux.

I understand how hal;Gy;AbbhyodIrghAt means "after hal and the GI Ap that have not been shortened". Might you please translate the other half of the rule, su;ti;sy;apRktaM hal, literally?

It means " su and the sip tip that turned into one consonant". They can only turn into one consonant by itazca.

Is the hal at the end of the rule useless?

Hey, I chewed down everything else for you. You figure out why that hal can't be taken out of the rule. That's your homework.

eG;hrasvAt sambuddheH ONPANINI 61069
After eG and short, (erase) calling.mmmmmmmmm ^ M+ C+ 524

Examples after eG --

sItA- + @calling →sambuddhauca sIte + s sIte "hey sItA"

kapi- + @calling →hrasvasyaguNaH kape + s kape "hey monkey"

guru- + @calling →hrasvasyaguNaH guro + s guro "hey teacher"

Examples after short --

rAma- + @calling → rAma "hey rAma"

nadI- + @calling →ambArtha nadi + s nadi "hey river goddess"

vadhU- + @calling →ambArtha vadhu + s vadhu "hey young lady"

Counterexample. subhrU- is not a nadI, but an uvaG, so it has long U (which is neither eG nor short) before the calling --

subhrU- + @calling → subhrUs "hey cutie"

See also rules that delete the calling .

Yet, kAlidAsa said subhru for "hey cutie" in the verse --

yadi tu tava samAgame tathaivamM prasarati subhru tataH kRtI bhaveyam "if the night could pass so slowly when I'm with you, cutie, I'd be happy"

Great poets break grammar sometimes. Doesn't mean you may do the same. You do that, I give you an F.

hrasvasya piti kRti tuk ONPANINI 61071
short gets tuk before pit kRt.mmmmmmmmm ^ M+ C+ 525

In morewords, the kRt affixes with p label, such as lyap kyap kvip, make the roots that end in a short vowel get t.

Examples with lyap --

ni + kR + lyap ni + kR + t + ya → nikRtya "humiliates and"

pra + stu + lyap prastutya "praises aloud and"

Examples with kyap --

iN' + kyap itya- "which has to be gone to"

stu + kyap stutya- "praiseworthy"

Examples with kvip --

agnim + ci + kvip agnicit- "fire-pileupper, arranger of the (sacrificial) fire"

indram + ji + kvip indrajit- "the one who defeated indra"

pustakam + bhR + kvip pustakabhRt- "book-carrier (name of sarasvatI)"

The kvip enders do not get any feminine affix. The kyap-enders get ajAdyataSTAp.

KAZIKA piti kRti parato hrasvAntasya dhAtoH tugAgamo bhavati. agnicit. somasut. prakRtya. prahRtya. upastutya. hrasvasya iti kim? AlUya. grAmaNIH. piti iti kim? kRtam. hRtam. kRti iti kim? paTutaraH. paTutamaH. grAmaNi brAhmaNakulam ityatra hrasvasya bahiraGgasya asiddhatvAt tug na bhavati.

Why do we bother to say kRt affixes?

Because taddhita affixes may have p too, but they don't make a short get t. So with tarap we get paTutara-, "sharper".

saMhitAyAm ONPANINI 61072
Not before pause.mmmmmmmmm ^ M- C+ 526

headline. The next rules (down to 61157) won't work before a pause.

Example. Rule ikoyaNaci below teaches "I to y before vowels". But because of this headline, this ikoyaNaci rule won't work on the I that is before a pause that in turn is before a vowel. So, when we join the words devI and uvAca to make a sentence, we have two options --

either we make no pause between the words and ikoyaNaci works --

devyuvAca "The Queen said."

or we make a pause and the ikoyaNaci does not work, even though I is before a vowel --

devI uvAca "The Queen. (pause). Said."

ikoyaNaci will work compulsorily when adding affixes --

kukkuTI- + au →ikoyaNaci kukkuTyau

because pausing midword is ungrammatical.

che ca ONPANINI 61073
( short gets tuk when in saMhitA) before ch.mmmmmmmmm ^ M+ C+ 527

We can reword this rule into --

" ch to cch after short "

Examples --

gam + laT tip → gam + zap + tip →iSugami gachati gatchati →stozzcu gacchati "goes"

chid + us →liTidhA cichidus citchidus →stozzcu cicchidus "they cut"

sas + chinatti →etattado sa + chinatti sat chinatti →stozzcu sac chinatti "he cuts"

atra + chAtraH atrat chAtraH →stozzcu atrac chAtraH "The student is here"

You may spell the last two as sa cchinatti and atra cchatraH if you like. But inria will expect you to type sacchinatti, atracchatraH.

This rule is quite useless nowadays, because everybody pronounces only cch, never ch, no matter if the spelling looks like ch or cch.

KAZIKA hrasvasya tukiti vartate. chakAre parataH saMhitAyAM viSaye hrasvasya tugAgamo bhavati. icchati. yacchati. hrasva eva atra AgamI, na tu tadantaH. tena cicchadatuH, cicchiduH ity atra tukabhyAsasya grahaNena na gRhyate iti halAdiHzeSeNa na nivartyate, nAvayavAvayavaH samudAyAvayavo bhavati iti.

AG;mAGoz ca ONPANINI 61074
AG mAG compulsorily (get tuk before ch).mmmmmmmmm ^ C+ 528

This rule teaches that the sound ch must be changed into the sound cch whenever it is after AG or mAG --

AG + chid + laT tip Ac + chid + tip Acchinatti "he steals"

mAG + chidaH mA cchidaH "don't cut!"

This is an exception to padAntAdvA, that says that after a wordfinal long vowel, the change is optional.

Most people make no difference between ch and cch and replace ALL ch with cch, so this rule is useless for them, and may be thought of as a spelling rule --

" after AG mAG, always spell the sound cch as cch"

KAZIKA tukiti anuvartate, che iti ca. AGo Gita ISadAdiSu caturSvartheSu vartamAnasya, mAGazca pratiSedhavacanasya chakAre paratastugAgamo bhavati. padAntAd veti vikalpe prApte nityaM tugAgamo bhavati. ISadarthe ISacchAyA AcchAyA. kriyAyoge AcchAdayati. maryAdAbhividhyoH A cchAyAyAH AccAyam. mAGaH khalvapi mA cchaitsIt. mA cchidat. GidviziSTagrahaNaM kim? AchAyA, AcchAyA. pramAchandaH, pramAcchandaH.

dIrghAt ONPANINI 61075
non- wordfinal long (compulsorily gets tuk before ch).mmmmmmmmm ^ C+ 529

Examples --

hrIcchati

mlecchati

Think of this rule as a spelling rule: " If the sound cch follows any vowel inside a word, spell it as cch ".

KAZIKA che tukiti vartate. dIrghAt paro yaH chakAraH tasmin pUrvasya tasya eva dIrghasya tugAgamo bhavati. hrIcchati. mlecchati. apacAcchAyate. vicAcchAyate.

The kAzikA shows the examples apacAcchAyate and vicAcchAyate, which look like yaG verbs to me, but no idea of what they mean. wafti.

pad%AntAd vA ONPANINI 61076
wordfinal ( long gets tuk before ch) optionally.mmmmmmmmm ^ C+ 530

What this rule is teaching us is that in pANini times, there were two different pronunciations of this example; some people used one and other people the other --

sA cchinatti "she cuts"

sA chinatti "she cuts"

Nowadays everybody pronounces ch and cch as cch, so this rule is useless, or is just a spelling rule --

" after a wordfinal long, spell the sound cch either as ch or as cch, you choose "

KAZIKA dIrghAt che tukiti vartate. padAntAd dIrghAt paro yaH chakAraH tasmin pUrvasya tasya eva dIrghasya pUrveNa nityaM prApto vA tug-Agamo bhavati. kuTIcchAyA, kuTIchAyA. kuvalIcchAyA, kuvalIchAyA. vizvajanAdInAM chadasi vA tugAgamo bhavati iti vaktavyam. vizvajanacchatram, vizvajanachatram. nacchAyAM karavo 'param. na chAyAM karavo 'param.

iko yaN aci ONPANINI 61077
ik to yaN before ac.mmmmmmmmm ^ M- C+ 531 sandhi

Before vowel, we replace i u R L, and their long versions, with y v r l respectively.

Examples with i I to y --

dadhi + atra dadhy atra "here's the curds"

kukkuTI + atra kukkuTy atra "here's the hen"

Examples with u U to v --

madhu + atra madhv atra "there's honey here"

vadhU- + jas vadhvas "young women"

Example with R to r --

pitR- + TA pitrA "with father"

There is an example of RR to r in rule zRRdRRprAM hrasvo vA --

zRRdRRpRR- + Am zRRdRRprAm "replace zRR dRR pRR with"

See also the exceptions akassa, ikosava, RtyakaH.

KAZIKA aci parataH iko yaN-Adezo bhavati. dadhyatra. madhvatra. kartrartham. lAkRtiH. ikaH pluta-pUrvasya savarna-dIrgha-bAdhanArthaM yaN-Adezo vaktavyaH. bho3i indram. bho3yindram. aci iti ca ayam adhikAraH saMprasAraNAc ceti yAvat.

eco 'y;av;Ay;AvaH ONPANINI 61078
ec to ay av Ay Av (before vowel).mmmmmmmmm ^ M- C+ 532 sandhi

The ec are e o ai au. So, before all vowels, we change --

e into ay

o into av

ai into Ay

au into Av.

This rule will always work as explained above inside words --

nI + laT sip → nI + zap + si →hardsoft ne + a + si nayasi "you lead"

bhU + laT tip → bhU + zap + ti →hardsoft bho + a + ti bhavati "is"

plu + laT ta → plu + zap + ta → plu + zap + te''' →hardsoft plo + a + te plavate "jumps"

It will also work on a wordfinal ec --

saha + nau + avatu saha nAv avatu "may he protect the two of us together"

zaranM tasmai + akSipam zaranM tasmAy akSipam "I shot him"

te + UcuH tay UcuH "they said"

And after that, rule lopazzA may work --

saha + nau + avatu saha nAv avatu saha nA:: avatu "may he protect the two of us together" (!!lopazzA

zaranM tasmai + akSipam zaranM tasmAy akSipam →lopazzA zaranM tasmA:: akSipam "I shot him"

te + UcuH tay UcuH →lopazzA ta:: UcuH "they said"

Yet, if the wordfinal is e o and the next word starts with a, this rule will not work, because of exception eGaHpa --

te + aplavanta →eGaHpa te 'plavanta

IMPORTANT. Even though pANini says that lopazzA is optional, it turns out that, because of an old custom, lopazzA is never applied IN WRITING to Av, and it is always applied to e o ai.

Yet, when talking Sanskrit, you may still apply it or not as you wish. Therefore, the examples tasmayakSipam, tayUcuH, sahanA::avatu above can reasonable be said to be mispellings, even though they are kosher.

KAZIKA ecaH sthAne aci parataH ayavAyAvityete AdezAH yathAsaGkhyam bhavanti. cayanam. lavanam. cAyakaH. lAvakaH. kayete. yayete. vAyAvavaruNaddhi.

vAnto yi pratyaye ONPANINI 61079
(o au to) av Av before y affix.mmmmmmmmm ^ M- C+ 533

Addition to ecoya, which replaces e ai o au before vowels. The o and the au will also be replaced before affixes that start with y.

Example before Nyat, a kRt affix --

yu + Nyat →acoJNiti yau + ya yAvya- "that should be excluded"

Example before Nya, a taddhita. The taddhita do not trigger ( acoJNiti, but do trigger taddhiteSva --

kuru- + Nya →taddhiteSva kauru- + ya →orguNaH kauro- + ya kauravya- "descendant of kuru"

See also orguNaH, gargAdi.

dhAtos tan-nimittasyaiva ONPANINI 61080
(o au) of root (to av Av before y -affix) only if that affix made the o au.mmmmmmmmm ^ C+ 534

Exception to vAntoyi above. If the o au belongs to a root, we replace it with av Av only before the y-affix that made the o au, like here --

lU + yat →hardsoft lo + yat lavya- "could be reaped"

but not before other y-affixes, like this yamAna here --

lU + yamAna + iJ →taddhiteSva lau + yamAna + i →yasyetica lauyamAni- "son of lUyamAna"

Here the U turned into au because of the iJ affix (see taddhiteSva), and not because of the yamAna affix. So we get no Av change before yamAna.

kSayya;jayyau zaky%Arthe ONPANINI 61081
kSayya- and jayya- only mean "-ible".mmmmmmmmm ^ C+ 535

These are built from kSi + yat and ji + yat. Only examples --

kSi + yat →hardsoft kSeya- kSayya- "destructible"

ji + yat → jeya- jayya- "conquerable"

When the meaning of yat is not "can be", this rule does not work --

kSi + yat →hardsoft kSeya- "should be destroyed"

ji + yat → jeya- "should be defeated"

KAZIKA kSi ji ityetayor dhAtvoH yati ratyaye parataH zakyArthe gamyamAne ekArasya ayAdezo nipAtyate. zakyaH kSetum kSayyaH. zakyo jetum jayyaH. zakyArthe iti kim? kSeyaM pApam. jeyo vRSalaH.