Rta:: ut ONPANINI 61111
R plus Gasi or Gas makes ur.mmmmmmmmm ^ C+ 577

Examples --

pitR + Gas pitur "of father"

kartR- + Gasi kartur "from maker"

KAZIKA GasiGasoH ityeva. RkArAntAd uttarayoH Gasi-GasoH ati parataH pUrvaparayoH ukAra ekAdezo bhavati. hoturAgacchati. hotuH svam. dvayoH SaSThI-nirdiSTayoH sthAne yaH sa labhate 'nyataravyapadezaM iti ur aN raparaH iti raparatvam atra kRtvA rAt sasya iti salopaH kartavyaH.

The literal translation of the rule is --

"R merges with the a of Gas Gasi into u"

Therefore, if you like splitting hairs, my translation above was sloppy, as this rule alone does not make pitur, as I said earlier. The rule just merges R + a into u and then two other rules finish the work --

pitR + Gas pitus →uraNraparaH piturs →saMyogAnta pitur "father's"

khya;tyAt parasya ONPANINI 61112
( Gas Gasi to u) after sakhi- and uncompounded pati-.mmmmmmmmm ^ C+ 578

Only examples:

sakhi- + Gas →ikoyaNaci sakhy + as sakhy + us →uraNraparaH sakhyurs →saMyogAnta sakhyur "of friend"

pati- + Gas patyur "of husband"

This rule won't work on compounded pati- --

senA- + pati- + Gas →gherGiti senApates "a general's"

KAZIKA GasiGasoH iti vartate, utiti ca. khyatyAtiti khizabdakhIzabdayoH tizabdatIzabdayozca kRtayaNAdezayoridaM grahaNaM, tAbhyAM parasya GasiGasoH ata ukArAdezo bhavati. sakhyurAgacchati. sakhyuH svam. patyurAgacchati. patyuH svam. khIzabdasya udAharaNam saha khena vartate iti sakhaH, tam icchati iti kyac sakhIyati. sakha yateH kvip sakhIH, tasya GasiGasoH sakhyuH iti. tIzabdasya api lUnam icchati lUnIyati, lUnIyateH kvipi lupte, lUnyurAgcchati. lUnyuH svam. niSThAnatvaM pUrvatrAsiddham 8-2-1 ityasiddham. vikRtanirdezAdeva iha na bhavati, atisakherAgacchati, senApaterAgacchati iti. sakhizabdasya kevalasya ghisaMjJA pratiSidhyate, na tadantasya.

Just in case you are curious: the nounbase "khya-tya" of this rule is short for "sakhy- and pathy-", that is, for "the sakhi pati that have turned into sakhy pathy before Gas Gasi". The i-enders only replace their i with y before Gas Gasi when gherGiti does not work, therefore only sakhi- and uncompounded pati- do that.

ato ror aplutAd aplute ONPANINI 61113
After a, ru to u (before a).mmmmmmmmm ^ M- C+ 579 sandhi

This rule only works after sasaju works, and after it works, AdguNaH and eGaHpa will always work. The effect of these four rules together is --

"replace wordfinal as plus the next a with o"

Example --

azvas + atra azvotra "horse is here"

Writing that in more detail --

azvas + atra →sasaju azva + ru + atra azva + u + atra →AdguNaH azvo + atra →eGaHpa azvotra

More examples --

rAmas + abravIt rAmobravIt "rAma said"

ziras- + adhinivAsini zirodhinivAsini "o peak-dwelleress"

See also Spelling of o before deleted a. .

KAZIKA ati, utiti vartate. akArAdaplutAduttarasya ro rephasya ukArAnubandhaviziSTasya akAre 'plute parata ukArAdezo bhavati. vRkSo 'tra. plakSo 'tra. bhobhagoaghoapUrvasya yo 'zi ityasmin prApte utvaM vidhIyate. rutvam api AzrayAt pUrvatrAsiddham iti asiddhaM na bhavati. ataH iti kim? agnir atra. tapara-karaNaM kim? vRkSA atra. sAnubandhagrahaNaM kim? svaratra. prAtaratra. ati ityeva, vRkSa iha. tasya api taparatvAdatra na bhavati. vRkSa AzritaH. aplutAd iti kim? susrotA3 atra nvasi. aplute iti kim? tiSThatu paya a3zvin. atra plutasya asiddhatvAtutvaM prApnoti iti aplutAdaplute iti ucyate.

hazi ca ONPANINI 61114
(After a, ru to u) before haz too.mmmmmmmmm ^ M+ C+ 580

When ru is after a and before haz, we replace that ru with u. This rule will only work when sasaju has replaced a wordfinal s with ru, and after this rule works, AdguNaH will always work. So the effect of the three rules combined is always --

" wordfinal as to o before haz "

Example before the haz letter d --

azvas + dravati azvo dravati "horse runs"

If you want the small details --

azvas + dravati →sasaju azva + ru + dravati azva + u + dravati →AdguNaH azvo dravati

More examples --

kapayas + nadyAm kapayo nadyAm "the monkeys are in the river"

manas- + bhis →svAdiSva manas ( word ) + bhis manobhiH "with minds"

KAZIKA hazi ca parataH ata uttarasya rorukArAdezo bhavati. puruSo yAti. puruSo hasati. puruSo dadAti.

sarvatra vibhASA goH ONPANINI 61122
(a stays) optionally after go-.mmmmmmmmm ^ C+ 581

Optional exception to eGaHpa.

Examples --

go_agram "(a country etc) whose main economic activity is bovine husbandry"

go_azvebhyas "from cattle and horses"

alternatively go'gram, go'zvebhyas. With eGaHpa.

KAZIKA sarvatra chandasi bhASAyAM ca ati parato goH eG prakRtyA bhavati vibhASA. go 'gram, go agram. chandasi apazavo vA anye goazvebhyaH pazavo goazvAn.

I spelled go'gram, go'zvebhyas above for clarity, but no one usese avagraha inside compounds. So, please remember to spell gogram, gozvebhyas.

pluta;pragRhyA:: aci nityam ONPANINI 61125
extralong and nonjoiner vowels stay before vowel.mmmmmmmmm ^ M+ C+ 582

After the extralong vowels and the nonjoiner vowels,

The rules that change a wordfinal vowel and the next vowel (for instance akassa, ikoyaNaci, eGaHpa, AdguNaH, ecoya) will not work if the first vowel is an extralong or a nonjoiner. Instead, we get a hiatus.

Example. Usually, because of eGaHpa, a disappears after a wordfinal e, such as the e of plavante here --

plavante + adya + kapayas + @pause → plavante 'dya kapayaH "today monkeys are jumping"

However, rule IdUdeddvi teaches that the final e of plavete is a nonjoiner. This makes eGaHpa not work --

plavete + adya + kapI → plavete adya kapI "today two monkeys are jumping"

KAZIKA plutAzca pragRhyAzca aci prakRtyA bhavanti. devadatta3atra nvasi. yajJadatta3idam Anaya. AzrayAdatra plutaH siddhaH. pragRhyAH agnI iti. vAyU iti. khaTve iti. mAle iti. aci ityanuvartamAne punarajgrahaNam Adezanimittasya aciH parigrahArtham. tena iha na bhavati, jAnu u asya rujati jAnvasya rujati. pragRhyAdukArAt parasya akArasya savarNadIrghatvaM pratyanimittatvAdatra prakRtibhAvo na bhavati. nityagrahaNam iha anuvartate. plutapragRhyANAM nityam ayam eva prakRtibhAvo yathA syAd, iko 'savarNe zAkalyasya hrasvaz ca 61127 ityetan mA bhUtiti.

iko 'savarNe zAkalyasya hrasvaz ca ONPANINI 61127
Optionally, ( wordfinal) ik stays and shortens before non- similar.mmmmmmmmm ^ C+ 584

Optional exception to ikoyaNaci.

So even though we usually apply ikoyaNaci in these --

dadhi + aicchat → dadhy aicchat "he wanted curds"

su-mahAMs tvat-sahAyeSu hary-RkSeSv asaMzayaH

the wordfinal i u may also stay --

dadhi + aicchat dadhi aicchat "he wanted curds"

su-mahAMs tvat-sahAyeSu hary-RkSeSu asaMzayaH

(This last line is from the rAmAyaNa. This option was taken mc.)

A long wordfinal must shorten --

kumArI + atra kumAri_atra "the princess is here"

KAZIKA iko 'savarne aci parataH zAkalyasya AcAryasya matena prakRtyA bhavanti, hrasvazca tasya ikaH sthAne bhavati. dadhi atra, dadhyatra. madhu atra, madhvatra. kumAri atra, kumAry atra. kizori atra, kizory atra. ikaH iti kim? khaTvendraH. asavarNe iti kim. kumArIndraH. zAkalyasya grahaNaM pUjArtham. ArambhasAmarthyAdeva hi yaNAdezena saha vikalpaH siddhaH. sinnityasamAsayoH zAkalapratiSedho vaktavyaH. siti ayaM te yonir RtviyaH. nityasamAse vyAkaranam. kumAryartham. ISA akSAdiSu chandasi prakRtibhAvamAtraM vaktavyam. ISA akSo hiraNyayaH. kA imare pizaGgilA. pathA agaman.

Rty akaH ONPANINI 61128
ak (optionally stays and shortens) before R.mmmmmmmmm ^ C+ 585

Ordinarily, because of AdguNaH, ikoyaNaci, akassa, we have --

a A + R → ar

i I + R → yR

u U + R → vR

R + R → RR

But this exception allows hiatus as well --

atra + RSayaH atra RSayaH "the sages are here"

bhavanti + RtavaH bhavanti RtavaH "there are seasons"

If the ak is long, it shortens --

kumArI + RSiz ca kumAri RSiz ca "the princess and the sage"

KAZIKA zAkalyasya hrasvazca ityetadanuvartate. RkAre parataH zAklyasya AcAryasya matena akaH prakRtyA bhavanti hrasvazca tasyakaH sthAne bhavati. khaTva RzyaH. mAla RzyaH. kumAri RzyaH. hotR RzyaH. Rti iti kim? khaTvendraH. akaH iti kim. vRkSAvRzyaH. savarNArthamanigarthaM ca vacanam.

diva::ut ONPANINI 61131
( wordfinal) div- to u.mmmmmmmmm ^ C+ 586

This div- is a f kvin-ender rootnoun that means "sky".

Examples with div made wordfinal by svAdiSva --

div- + bhis diu + bhis →ikoyaNaci dyubhis "with skies"

div- + sup' diu + su →ikoyaNaci dyu + su →kric dyuSu "in the skies"

Examples with div made wordfinal by supodhA --

divam + spRzati "touches the sky" → div- + spRz + su dyu + spRz + su →halGyA dyuspRz →kvinpra dyuspRk "skyscraper" ( anything that touches the sky )

div- + pati + su dyu + pati + su → dyupatis "skylord" ( the Sun )

When the v is not wordfinal this rule won't work, so we say divau, divas.

If the su is strong, this rule won't work (because the su is strong, the v is not made wordfinal by svAdiSva), and we get dyaus (by diva::aut).

But in the end of a compound like vimaladiv- "that has clear skies", when it is neuter, the su will be erased by svamor, and this rule works, and we say --

vimaladyu dinam "a day with clear sky"

KAZIKA eGaH padAntAdati ity ataH padagrahaNam anuvartate. divaH iti prAtipadikaM gRhyate, na dhAtuH, sAnubandhakatvAt. divaH padasya ukArAdezo bhavati. divi kAmo yasya dyukAmaH. dyumAn. vimaladyu dinam. dyubhyAm. dyubhiH. niranubandhakagrahaNAdiha na bhavati, akSadyUbhyAm, akSadyUbhiH iti. taparakaranam UTho nivRttyartham, dyubhyAm, dyubhiH iti. atra hi paratvAt UTH prApnoti. padasya iti kim? divau. divaH.

etat;tadoH su-lopo 'kor a-naJ-samAse hali ONPANINI 61132
Delete su of sas eSas before consonant.mmmmmmmmm ^ M+ C+ 587

eSas and sas mean "this" and "that". They are the masculine pronouns tad- and etad- with su added --

azvo mayA dRSTasH saH "I saw that horse"

azvo mayA dRSTa eSaH "I saw this horse"

If they are before a consonant, in saMhitA, they lose their final su, turning into sa and eSa --

azvo mayA sa dRSTaH "I saw that horse"

azva eSa mayA dRSTaH "I saw this horse"

Before vowel, this rule does not delete the su, but other rules delete or change it anyway --

eSas + azvaH →atoro eSo 'zvaH "this horse"

eSas + uSTraH →lopazzA eSa_uSTraH "this camel"

So, the su only manages to survive when it is before pause --

eSas + @pause →sasaju eSar →kharava eSaH "this one"

KAZIKA etat;tadau yAvakakArau naJsamAse na vartate tayor yaH suzabdaH, kazca tayoH suzabdaH? yaH tadarthena sambaddhaH, tasya saMhitAyAM viSaye hali parato lopo bhavati. eSa dadAti. sa dadAti. eSa bhuGkte. sa bhuGkte. etattadoH iti kiM? yo dadAti. yo bhuGkte. sugrahaNaM kim? etau gAvau carataH. akoH iti kim? eSako dadAti. sako dadAti. tanmadhyapatitas tadgrahaNena gRhyate iti rUpabhede 'pi sAkackAv etat-tadAv eva bhavataH. a-naJ-samAse iti kim? aneSo dadAti. aso dadAti. uttarapad%Artha- pradhAnatvAn naJ-samAsasya etattadoreva atra sambaddhaH suzabdaH. hali iti kim? eSo 'tra so 'tra.

The akos and anaJsamAse of the original rule clarify that this rule won't chop off the su of the following four words, even tough they are built from sa eSa --

eSakas "this very one"

sakas "that very one"

asas "someone else, not that one"

aneSas "someone else, not this one"

I have never heard those words or seen them written outside of the commentaries to this rule.

sam;pary;upebhyaH karotau bhUSaNe ONPANINI 61137
After sam pari upa, when meaning decorating, kR gets s(uT).mmmmmmmmm ^ C+ 588

This rule will replace the kR that means decorating with skR.

Examples --

upa + kR + tRc upa + skR + tR upaskartR- "decorator"

pari + kR + tRc pari + skR + tR pariskartR- →parinivi pariSkartR- "decorator"

This rule won't work when the meaning is not "decorating" --

upa + kR + tRc upakartR- "helper, furnisher"

KAZIKA sam pari upa ity etebhyaH bhuSaN%Arthe karotau parataH suT kAt pUrvo bhavati. saMskartA. saMskartum. saMskartavyam. atra saMpuMkAnAM satvam iti samo makArasya sakAraH, pUrvasya cAkArasya anunAsikaH. pariSkartA. pariSkartum. pariSkartavyam. suTstusvaJjAm iti Satvam. upaskartA. upaskartum. upaskartavyam. bhUSaNe iti kim? upakaroti. sampUrvasya kvacidabhUSaNe 'pi suDiSyate, saMskRtamannam iti.

citaH ONPANINI 61163
cit enders get acute on the last.mmmmmmmmm ^ C+ 590

When an affix is cit, the last vowel of the result of adding stem and affix gets the acute. As in --

uccais + akac uccakais ( accent on the ai )

kR + liT jha → cakR + irec cakrire ( accent on the e )

kR + tRc kartR- ( accent on the R )

There is another kartR- made with tRn --

kR + tRn → kartR- ( with accent on the a by JnityAdi )

Back to labels and accent .

kitaH ONPANINI 61165
kit ( taddhita -ender has acute on the end).mmmmmmmmm ^ C+ 591

For instance, the kRt affix kta is kit. So in --

nI + kta → nItaH "was led"

the accent is on the taH.

Back to labels and accent .

KAZIKA taddhitasya ityeva. taddhitasya kito 'nta udAtto bhavati. naD´AdibhyaH phaG nADAyanaH. cArAyaNaH. prAg vahateS Thag AkSikaH. zAlAkikaH.

aSTano dIrghAt ONPANINI 61172
( weak is acute) after long aSTan-mmmmmmmmm ^ C+ 592 accent

aSTan has a long at the end only when rule aSTana:: A vibhaktau worked, changing it into aSTA-. Therefore --

aSTAbhiH

aSTAbhyaH

aSTAsu

have acute on the last vowel.

KAZIKA aSTano dIrghAntAd a-sarvanAmasthAna-vibhaktir udAttA bhavati. aSTAbhiH. aSTAbhyaH. aSTAsu. ghRtAdipAThAtaSTanzabdo 'nt%odAttaH, tatra jhaly upottamam 61180 ityasya apavAdo vibhaktir eva udAtta-tvaM vidhIyate. dIrghAd iti kim? aSTasu prakrameSu brAhmaNo 'gnIn AdadhIta. idam eva dIrgha-grahaNam aSTana At-tva-vikalpaM jJApayati, kRtAtvasya ca SaT-saMjJAM jJApayati. anyathA hyAtvapakSe sAvakAzo 'STanaH. svaraH paratvAdanAtvapakSe SaTsvareNa bAdhiSyate iti kim dIrgha-grahaNena.

We know that the rule aSTana:: A vibhaktau is optional because this rule contains the word dIrghAt, to mean "after the aSTan that ends in a long". Consider: if rule aSTana:: A vibhaktau were compulsory, then all aSTan- before a weak would become aSTA, and specifying dIrghAt here would be redundant.

tit svaritam ONPANINI 61185
An affix with t label has falling on its firstmmmmmmmmm ^ C+ 593

Exception to AdyudAttazca, that makes affixes acute in the first by default.

So --

kR + Nyat + su → kAryam "(that) has to be done"

has falling on the ya.

Back to labels and accent .

liti ONPANINI 61193
The syllable before a lit affix gets the acute.mmmmmmmmm ^ C+ 594

Exception to AdyudAttazca.

lit means "what has label ell" (not same as liT, a past tense).

Examples with the lit affixes are (N)a(l), (N)vu(l) --

kR + liT tip → kR + Nal cakAra "he made"

with accent on the kA

kR + liT mip → kR + Nal cakara "I made"

with accent on the ka

kR + san + Nvul cikIrSaka- "one that is about to make"

with accent on the kIr

Back to labels and accent .

KAZIKA liti pratyayAt pUrvam udAttaM bhavati. cikIrSakaH. jihIrSakaH. bhaurikividham. bhaulikividham. aiSukAribhaktam.

Jit is what has label Jmmmmmmmmm ^ M- C+ 595

Some Jit affixes:

aJ ghaJ iJ JyaG vuJ naJ

Rule svaritaJi says Jit roots, like kR and UrNu, are flattybendy.

nit is what has label n.mmmmmmmmm ^ C+ 596

The nit affixes put the accent at the start (by JnityA), just like the Jit affixes.

Some nit affixes --

zyan

tumun

san

ktin

kvin

Than

tRn

vun

Jnity Adir nityam ONPANINI 61197
The start before Jnit (is acute).mmmmmmmmm ^ M- C+ 597 accent

The Jit affixes and the nit affixes place the accent on the first vowel of whatever they are added to. HYPER-CAREFUL here please -- I said nit, NOT Nit.

Examples. zyan, vun and yaJ are Jnit affixes. That's why this yudhyate has the accent on the yu by this rule --

yudh + kartari laT ta → yudh + zyan + ta → yudhya + te''' → yudhyate "he's fighting"

While this other yudhyate has the accent on the yak (by Ady-udAttazca) --

yudh + akartari laT ta → yudh + yak + ta → ydhya + te''' → yudhyate "fighting is going on"

KAZIKA Jiti niti ca nityam Adir udAtto bhavati. gargAdibhyo yaJ gArgyaH. vAtsyaH. vAsudev%ArjunAbhyAM vun vAsudevakaH. arjunakaH. pratyayasvar%ApavAdo 'yaM yogaH. pratyayalakSaNam atra na iSyate, tena gargAH, bidAH, caJcAH ity atra yaJi kani ca lupte na bhavati.

If you find Jni hard to chant, just chant Jini instead. Like the loudspeaker icon at panini research tool 6-1-197.

yato '-nAvaH ONPANINI 61213
(Two-vowel) yat-enders (accent the first), except after nau-.mmmmmmmmm ^ C+ 598 accent

Exception to titsvaritam. Roots with yat, and nouns with yat'', get the acute before the affix.

Example --

ji + yat →hardsoft je + ya → jeya- "defeatable" ( accent on je )

Counterexample --

nau + yat'' →vAntoyi nAv + ya → nAvya- "suitable for ships, navigable" ( accent on vyam )

Back to labels and accent .

KAZIKA niSthA ca dvyajanAt 61205. ity ato dvy-aj-grahaNam anuvartate. yat pratyayAntasya dvyaca AdirudAtto bhavati na cen@ nau-zabdAt paro bhavati. aco yac ceyam. jeyam. zarIrAvayavAd yat kaNThyam. oSThyam. tit svaritam ity asya apavAdaH. a-nAvaH iti kim? nAvyam. dvy-acaH ity eva, cikIrSyam. lalATyam.

upottamaM riti ONPANINI 61217
rit affixes have acute in the next-to-last vowel.mmmmmmmmm ^ C+ 599

Exception to AdyudAttazca.

A rit is whatever has label r. Examples with the rit affixes anIya(r) kelima(r) --

kR + anIyar → karaNIya- ( accent on NI )

pac + kelimar → pacelima- ( accent on li )

Back to labels and accent .

bahuvrIhau prakRtyA pUrva-padam ONPANINI 62001
In a longhorn, the former keeps its accent.mmmmmmmmm ^ C+ 600

So the latter loses it.

This is not a vedic rule, but in practice it might as well be -- I want to remind the reader that nowadays accent rules are only applied when reciting the veda . So that you won't lose sleep because of the horror story below.

Traditional example (translated into English) --

The word "indra-kíller" is a tatpuruSa compound, and means someone who will kill indra. The word indra lost its accent when compounded.

But "índra-killer" is a longhorn compound, and means someone whose killer is indra. Namely, someone who was or will be killed by indra. The word for killer lost its accent when compounded.

There is an old story about tvaSTR-, who asked brahmA to get an indrakiller son, but, when asking, said "índra-killer" instead of "indra-kíller". So brahmA did not give him what he wanted, but what he had asked for. Gods use REAL Sanskrit, and not the botched debased accentless dialect that passes for it in this wretched kaliyuga.

Yet, when you are not talking to gods, nor chanting the veda, NVM about accents.

BTW, I found that story about tvaSTA in a blog, hope it's still there --

tvaSTR mispronounces indrazatru

KAZIKA pUrvapada-grahaNaM atra pUrvapada-sthe svare udAtte svarite vA vartate. bahuvrIhau samAse pUrva-padasya yaH svaraH sa prakRtyA bhavati, sva-bhAvenAvatiSThate, na vikAram anudAtta-tvam Apadyate. samAsAntodAttatve hi sati anudAttaM padam ekavarjam 61158 iti so 'nudAttaH syAtiti samAsAntodAttatvApavAdo 'yam Arabhyate. kArSNottarAsaGgAH. kRSNo mRgaH tasya vikAraH kArSNaH, prANirajatAdibhyo 'J iti aJpratyayAnto Ji-svareNa Ady-udAttaH. yUpavalajaH. yUpa-zabdaH uNAdiSu kusuyubhyazca iti papratyayAntaH. tatra ca dIrghaH iti niditi ca vartate tena AdyudAttaH. brahmacAripariskandaH. brahamacArizabdaH kRd-uttarapada-prakRti-svareNa antodAttaH. snAtaka-putraH. snAtaka-zabdaH kan-pratyay%Anto nit-svareNa Ady-udAttaH. adhyApaka-putraH. lit-svareNa adhyApaka-zabdo madhy%odAttaH. zrotriya-putraH. zrotriya-zabdo nitvAd AdyudAttaH. manuSya-nAthaH. manuSya-zabdaH tit svaritam iti svarit%AntaH. udAtta-grahaNam svarita-grahaNaM cAtra anuvartate, tena sarvAnudAtte pUrvapade vidhireva na asti iti samAsAntodAttatvaM bhavati. samAbhyAgaH iti samazabdo hi sarvAnudAtaH.

hal;ad-antAt saptamyAH saMjJAyAm ONPANINI 63009
After consonant or a, seventh stays, if tag.mmmmmmmmm ^ C+ 601

That is to say, supodhA does not work on a tag whose former is a nounbase that ends in a or consonant, and got seventh.

For instance, the nounbase of araNye "in the forest" is araNya-, ending in a --

araNya- + Gi →AdguNaH araNye "in the wild"

Therefore when this araNye is former in a tag, supodhA does not work --

araNyetilakAH "disappointing things"

But if araNye tilakAH is used in its literal meaning of "wild sesame plants", then the compounding is not allowed --

araNye tilakAH "there are sesame plants in the forest"

See gaviyudhibhyAM sthiraH for another example.

KAZIKA halantAd adantAc ca uttarasyAH saptamyAH saMjJAyAm alug bhavati. yudhiSThiraH. tvacisAraH. gaviSThiraH ity atra tu gaviyudhibhyAM sthiraH 83095 ity ata eva vacanAd aluk. adantAt araNye-tilakAH. araNye-mASakAH. vane-kiMzukAH. vaneharidrakAH. vane-balbajakAH. purvAhNe-sphoTakAH. kUpe-pizAcakAH. haladantAditi kim? nadyAM kukkuTikA nadIkukkuTikA. bhUmyAM pAzAH bhUmipAzAH. saMjJAyAm iti kim? akSazauNDaH. hRddyubhAM GeH. hRd divityetebhyAm uttarasya Ger alug bhavati. hRdispRk. divispRk.

Incidentally. When I was learning this rule for the first time I wondered why pANini wanted araNyetilakaH to be a compound if it had one meaning and not a compound if it had the other meaning, as compound and not compound where going to be pronounced the same anyway. Boy I was so ignorant back then. That doubt of mine had at least three misunderstandings in it.

AnaG Rto dvandve ONPANINI 63025
Replace R -enders with An(aG) (when former) in a dvandva.mmmmmmmmm ^ M- C+ 602

Example:

mAtR + pitR- + au mAtAn + pitR- + au →nalopaHprA mAtA + pitR- + au → mAtA;pitarau "parents"

KAZIKA RkArAntAnAM vidyAyonisambhandhavAcinAM yo dvandvas tatra uttarapade pUrvapadasya AnaGAdezo bhavati. hotApotArau. neSTodgAtArau. prazAstApratihartArau. yonisambandhebhyaH mAtApitarau. yAtAnanAndarau. makAroccAraNaM raparatvanivRttyartham. RtaH iti kim? pitRpitAmahau. putre ityatra anuvartate, RtaH iti ca. tena putrazabde 'pyuttarapade RkArAntasya anaGAdezo bhavati. pitAputrau. mAtAputrau.

An@ mahataH samAnAdhikaraNa-jAtIyayoH ONPANINI 63046
mahat- to A before samereferent or jAtIya.mmmmmmmmm ^ C+ 603

So mahat- becomes mahA- when former in a karmadhAraya --

mahat- + rAjan- mahA- + rAjan- →rAjAhassa mahArAja- "great king"

or a longhorn --

mahAntau bahU asya →supodhA mahat- + bAhu- mahAbAhu- "big-armed"

or before jAtIya "sort of" --

mahAjAtIya- "sort of big, biggish, moderately big"

But this rule won't work on

mahataH putraH →SaSThI mahatputra- "son of a big man"

because here the two halves of the compound are not samereferent, as mahat- means the father and putra- means the son.

KAZIKA samAnAdhikaraNe uttarapade jAtIye ca pratyaye parato mahataH AkArAdezo bhavati. mahAdevaH. mahAbrAhmaNaH. mahAbAhuH. mahAbalaH. jAtIye mahAjAtIyaH. samAnAdhikaraNajAtIyayoH iti kim? mahataH putraH mahatputraH. lakSaNoktatvAd eva atra na bhaviSyati iti ced bahuvrIhAv api na syAd mahAbAhuH iti. tadarthaM samAnAdhikaraNagrahaNaM vaktavyam. amahAn mahAn sampanno mahadbhUtazcandramAH ityatra gauNatvAn mahadarthasya na bhavatyAtvam. mahadAtve ghAsakaraviziSTeSu upasaGkhyAnaM puMvadvacanaM ca asamAnAdhikaraNArtham. mahatyAH ghAsaH mahAghAsaH. mahatyAH karaH mahAkaraH. mahatyAH viziSTaH mahAviziSTaH. aSTanaH kapAle haviSyupasaGkhyAnam. aSTakapAlaM caruM nirvapet. haviSi iti kim? aSTakapAlaM brAhmaNasya. gavi ca yukte 'STana upasaGkhyAnaM kartavyam. aSTAgavena zakaTena. yukte iti kim? aSTagavaM brAhmaNasya taparakaraNaM vispaSTArtham.

arur;dviSad;aj-antasya mum ONPANINI 63067
arus-, dviSat-, and vowel-enders get mum (when they are former before a khit-ender).mmmmmmmmm ^ C+ 605

Some compounds (such as janamejaya-, paranMtapa-, arunMtuda-, parapuraJMjaya-) get m in the middle. This happens when the latter was made with a khit affix ( khaz or khac), and the former is arus-, or dviSat- (the one from dviSomitre), or ends in a vowel.

Examples of khaz and khac enders --

ej + Nic + khaz → eji + khaz → eji + zap + khaz →hardsoft eje + a + khaz →ecoya ejaya + khaz →atoguNe ejaya "that makes (them) tremble"

tap + khac → tapa

ji + khac →hardsoft je + a →ecoya jaya

All of these must compound with their object. The object gets mum by this rule. Example with a vowel-ender, para- "enemy" --

para- + tap + khac paramtapa- →monusvA paraMtapa →anusvA paranMtapa- "enemy-scorcher"

examples with dviSat- and arus-, thatalso mean "enemy" --

dviSat- + tap + khac dviSamt tapa- →saMyogAnta dviSamtapa- →monusvA dviSaMtapa →anusvA dviSanMtapa- "enemy-scorcher"

arus- + tud + khaz arumstuda →saMyogAnta arumtuda →monusvA aruMtuda →anusvA arunMtuda- "enemy-beater"

janamejaya- "he that makes people tremble (pn of a king)"

nitambamejayA- "she that twerks vigorously"

parapuraJMjaya- "conqueror of the cities of the enemy"

KAZIKA arus dviSad ity etayor ajanatAnAM ca khid-anta uttara-pade mum-Agamo bhavati anavyayasya. aruntudaH. dviSantapaH. ajantAnAm kAlimmanyA. arurdviSadajantasya iti kim? vidvanmanyaH. anavyayasya ityeva, doSAmanyamahaH. divAmanyA rAtriH. antagrahaNaM kim? kRtAjantakArya-pratipatty-artham. ato hrasve kRte mum bhavati.

na-lopo naJaH ONPANINI 63073
Delete the n of naJ.mmmmmmmmm ^ M+ C+ 606

na(J) means "no" or "lacking". It is used only as a former.

The J of naJ is a label, and the n is deleted by this rule, so only a stays. This "a" is the same negative element that appears in words like a-theist, a-scorbic, an-hidrid.

In a tatpuruSa --

naJ + brahmaNaH → abrahmaNaH "fake brahmin, he's not a brahmin"

In a longhorn --

naJ + brahmaNaH → abrahmaNaH "(a country) devoid of brahmins"

If what is after naJ starts with a vowel, next rule tasmAnnuDaci will work.

KAZIKA naJo nakArasya lopo bhavati uttarapade. abrAhmaNaH. avRSalaH. asurApaH. asomapaH. naJo nalope 'vakSepe tiGyupasaGkhyAnaM kartavyam. apacasi tvaM jAlma. akaroSi tvaM jAlma.

tasmAn@ nuD aci ONPANINI 63074
vowel after naJ gets nuT.mmmmmmmmm ^ M+ C+ 607

Addition to naloponaJaH. When naJ is before a vowel-starter, that vowel-starter gets nuT. In other words, we add n in front of the vowel-starter, this way --

Example. Rtam starts with a vowel, so it gets nuT here --

naJ + Rtam "true" →naloponaJaH a + Rtam a + (Rtam + nuT) →Adyantau a + nRtam → anRtam "untrue, an untruth"

Notice that this rule adds n in front of the Rtam, and not at the end of the naJ. It must be done that way because otherwise rule Gamohra would screw things up (or maybe nalopaHprA would, I'm not sure). Anyway, spelling anRtam is fine, spelling an-Rtam is fine, and I avoid a-nRtam because it confuses my students big time.

KAZIKA tasmAl lupta-nakArAn naJaH nuT-Agamo bhavati aj-AdAv uttarapade. anajaH. anazvaH. tasmAd iti kim? naJa eva hi syAt. pUrvAnte hi Gamo hrasvAd aci GamuN nityam iti prApnoti.

v'' .opasarjanasya ONPANINI 63082
( saha- to sa-) optionally in a longhorn.mmmmmmmmm ^ C+ 608

Example --

Compounding saha- with putrais we get sahaputra-, that means "accompanied with sons".

sahaputra- is a longhorn because rules zeSobahuvrIhiH and tenasahetitulyayoge say so. Therefore this rule applies --

saha- + putrais sa- + putrais → saputra- "accompanied with sons"

Notice that the bhis ending of putrais is lost. This happens because saputra- will be used asan adjective, and will lose its own gender and case and number to get the gender and case and number of whatever noun it will describe --

pazyAmi pitRRn saputrAn "I see fathers with sons"

pazyAmi mAtRRsH saputrAH "I see mothers with sons"

This rule is optional --

saha- + putrais → sahaputra- "accompanied with sons"

This rule will not work on non- longhorn compounds like sahayudhvan- "one who fights together, a brother-in-arms", which is not a longhorn because it means a particular sort of yudhvan- "warrior".

KAZIKA upasarjana-sarvAvayavaH samAsaH upasarjanam. yasya sarve 'vayavA upasarjanIbhUtAH sa sarvopasarjano bahuvrIhir gRhyate. tadavayavasya sahazabdasya vA sa ityayam Adezo bhavati. saputraH, sahaputraH. sacchAtraH, sahacchAtraH. upasarjanasya iti kim? sahayudhvA. sahakRtvA. sahakRtvapriyaH, priyasahakRtvA iti iha bahuvrIhau yaduttarapadaM tat paraH sahazabdo na bhavati iti sabhAvo na bhavati.