71027 yuSmad and asmad ←

chunk 38: 71078 /zatR with /num

→ 72009 /iT rules

71078 zatR doesn't after stammered. nAbhyastAcchatuH
71079 But optionally before zi. vAnapuMsakasya
71080 After "A a", gets num before zI GI. AcchInadyornum
71081 Compulsorily if there is zap zyan. zapzyanornityam
71082 anaDuh- before su. sAvanaDuhaH
71084 div- to au . divaaut
71085 pathin- mathin- RbhukSin- to A before su. pathimathyRbhukSAmAt
71086 i to a before strong. itotsarvanAmasthAne
71087 th to nth . thonthaH
71088 Tail of wimpy to lopa. bhasyaTerlopaH
71089 pums- to as . puMsosuG
71090 Nit after go-. gotoNit
71091 First person Nal optionally. NaluttamovA
71092 Non-calling after sakhi-. sakhyurasambuddhau
71093 to anaG before su. anaGsau
71094 Replace R-enders, uzanas-, purudaMsas- and anehas- . RduzanaspurudaMsonehasAJca
71098 catur- and anaDuh- get Am' . caturanaDuhorAmudAttaH
71100 RR of root to ir. RRta::iddhAtoH
71101 nexttolast RR to ir. upadhAyAzca
71102 ur after root labial. udoSThyapUrvasya
72001 ik to vRddhi before sic that is before flat. sicivRddhiHparasmaipadeSu
72002 a of r-l-ender root to vRddhi . atorlAntasya
72003 vad vraj and hal-enders to . vadavrajahalantasyAcaH
72004 But not before iSic. neTi
72007 Light a of a consonant-starter . atohalAderlaghoH


n/AbhyastAc chatuH ONPANINI 71078
zatR doesn't (get num) after stammered.mmmmmmmmm glosses glosses ^ M- C+ 716 num

(Exception to ugidacA, that would otherwise num all zatR before strong.)

Example with a stammered from jakSi --

daridrA + zatR m + su znAbhyastayorAtaH daridr + at + s halGyA daridrat "that is poor, sad, suffering "

Example with a stammered made by zlu --

dA + zlu + zatR zlau dadA + at znAbhya dadat- "giving"

dadat- f + amdadat + GI + am amipUrv dadatIm f "that gives"

dadat- m + amdadatam m "that gives"

dadat- m + su halGyA dadat m "that gives"

As in --

pazyAmi rAjAnanM dadataGM gAH "I see that the king is giving cows"

This masculine dadat sounds, misleadingly, like the neuter --

dadat- n + su OR au svamorna dadat n "that gives"

As in --

pazyAmi rakSo dadad gAH "I see that the demon is giving cows"

See also exception vAnapuMsakasya right below.

See also zatR getting num before zI GI. .


vA napuMsakasya ONPANINI 71079
But optionally before zi.mmmmmmmmm glosses glosses ^ M- C+ 717 num

After stammered, zatR gets num optionally before zi, in spite of nAbhyastAcchatuH that would forbid num.

Examples --

dadanti kulAni "families that give"

dadati kulAni "families that give"

jakSanti kulAni "families that eat "

jakSati kulAni "families that eat "

Careful. In inria, these are the blue green dadati jakSati, not the red ones!

See also zatR getting num before zI GI. .


Ac chI;nadyor num ONPANINI 71080
After "A a", ( zatR optionally) gets num before zI GI.mmmmmmmmm glosses glosses ^ M- C+ 718 num

As GI and zI are not strong, usually they would not cause zatR to get num. Yet this rule says that they get num optionally.

Yet, because of exceptions ( zapzyanornityam, nAbhyastAcchatuH), in fact we only get this optionality after za, after the sya made by lRTassa, and after the lukclass roots that end in A, such as yA.

Examples after za --

before GI --

tud + za + zatR + GI atoguNe tudatI → tudantI- f "that slaps, that vexes, that bullies"

tudantI brAhmaNI "a brAhmaNI that bullies"

before zI --

tudantI kule "two bullying families"

Example after sya --

kumbhaGM kariSyantI brAhmaNI "a brAhmaNI that is about to make a pot"

kumbhAn kariSyantI kule "two families that are about to enter the ceramics business"

Example after yA --

yAntI brAhmaNI "a brAhMaNI that goes"

yAntI kule "two families that go"

If we don't use this rule, we get the same forms without the num -- tudatI brAhmaNI, tudatI kule, yAtI brAhmaNI etc.

Back to zatR getting num before zI GI. .

Notice that pANini uses the word zInadyos to mean "before zI and GI". I guess he didn't say zIGyos because that might mean "before zI GI Gi".


zap;zyanor nityam ONPANINI 71081
Compulsorily if there is zap zyan.mmmmmmmmm glosses glosses ^ M- C+ 719

Exception to the previous rule, AcchI. When there is zap or zyan before the zatR that is before zI GI, then the num is compulsory.

Examples with GI after a feminine --

pac + zatR + GI + supacatIs → pacantIs halGyA pacantI f "that cooks"

pacantI brAhmaNI "a cooking brAhmaNI"

dIvyantI brAhmaNI "a gambling brAhmaNI"

Examples with zI after a neuter --

pacantI rakSasI "two cooking male demons"

dIvyantI rakSasI "two gambling male demons"

So num is compulsory after most roots, because the roots that get zap zyan before zatR make up about 80 percent of all roots -- all the zapclass, zyanclass, Nichclass and causative roots.

See also zatR getting num before zI GI. .


sAv anaDuhaH ONPANINI 71082
anaDuh- (gets num) before su.mmmmmmmmm glosses glosses ^ M+ C+ 720 num

The word anaDuh-, even though it is most often used to refer to bulls and oxen, properly means cart-puller. It used to be a compound of anas- "cart", anomalously changed to anaD, and vah, a rootnoun for "dragger". With stretching by vacisvapi.

Only example --

anaDuh- m + su caturana anaDuh + Am' + su midaco anaDuAh + s → anaDuAnhs ikoyaNaci anadvAnhs halGyA anadvAnh saMyogAntasya anaDvAn "ox, bull"

By a vArttika, this rule won't work when anaDuh- means a cart-pulling female. In that case, there is GI, and caturana is optional. So with the A from caturana we get --

anaDuh- + GI + su caturana anaDvAh + GI + s halGyA anaDvAhI "cart-pulling cow"

and without it --

anaDuh- + GI + su halGyA anaDuhI "cart-pulling cow"


diva::aut ONPANINI 71084
div- to au (before su).mmmmmmmmm glosses glosses ^ C+ 721 sup

This div- is a nounbase that means "sky", and has nothing to do with the root div "gamble".

Only example --

div- f + su → di + au + s ikoyaNaci dyaus "sky"

Same as with naus gaus, no rule erases the calling after au. So, whenever you address the sky or its goddess, start with he dyauH.

See also diva::ut.


pathi;mathy;RbhukSAm At ONPANINI 71085
pathin- mathin- RbhukSin- to A before su.mmmmmmmmm glosses glosses ^ C- 722 sup

There are exactly three examples:

pathin- m + su itotsarva pathan- + su thonthaH panthan- + su → pantha- + A + su akassa panthA- + supanthAs "way, road"

manthAs "churning stick"

RbhukSAs "pn (of indra)"

This A being long, eGhra won't work --

he rbhukSAH "hey indra"

Back to pathin-.


ito 't sarvanAmasthAne ONPANINI 71086
i (of pathin- mathin- RbhukSin-) to a before strong.mmmmmmmmm glosses glosses ^ C+ 723

Examples --

pathin- + au thonthaH panthin + au → panthan- + au rAjAnam panthAnau "two roads"

mathin- + am → .. → manthAnam "churning stick"

supanthAni rAjyAni "kingdoms that have a good transportation infrastructure"

Back to pathin-.


tho 'nthaH ONPANINI 71087
th (of pathin- mathin-) to nth (before strong).mmmmmmmmm glosses glosses ^ C+ 724 sup

Example --

pathin- + au → panthin + au itotsarva panthan + au rAjAnam panthAnau "two roads"

Back to pathin-.


bhasya Ter lopaH ONPANINI 71088
Tail of wimpy ( pathin- mathin- RbhukSin-) to lopa.mmmmmmmmm glosses glosses ^ C+ 725 sup

Examples --

pathin- + TA → path + TApathA "by road"

pathin- + Gi → path + Gipathi "at the road"

Back to pathin-.


puMso 'suG ONPANINI 71089
pums- to as(uG) (before strong).mmmmmmmmm glosses glosses ^ C+ 726 sup

The nounbase pums- means "man" or "male". This rule says "replace it with as(uG)". The G label is for Gicca, and the u makes ugidacA work before all strong. Then sAnta works before all strong except the calling --

pums + su → pumas (u) + s ugidacA pumanss sAnta pumAnss halGyA pumAns saMyogAnta pumAn "man"

pums + au → pumas (u) + au ugidacA pumansau sAnta pumAnsau nazcApa pumAMsau "two men"

pums + jas → .. → pumAMsas

pums + am → .. → pumAMsam

pums + @calling → pumans halGyA puman "hey man"

When wimpy --

pums + zas nazcApa puMsas

pums + TA nazcApa puMsA

When word --

pums + sup' saMyogAnta pum + su monusvA puMsu

and before bh, anusvA is optional --

pums + bhis saMyogAnta pum + bhis monus puMbhis

pums + bhis saMyogAnta pum + bhis monus puMbhis anusvA pumbhis

As the m is now technically wordfinal, the CRAZY SPELLING RULE works, and these will usually be spelled pumMbhis, with a topdot on the pu. But don't panic if you see pumbhis or even pu~bhis sometimes. Don't worry about that mess, pronounce always pumbhiH, and spell however you like.


goto Nit ONPANINI 71090
( Strong is) Nit after go-.mmmmmmmmm glosses glosses ^ M- C+ 727 sup

There are just three examples, su, au and jas --

go- + su → go + s(uN) acoJNiti gau + sugaus "cow"

go- + @calling → .. → gaus "hey cow" ( same steps )

go- + au → .. → gau + au ecoya gAvau "two cows"

go- + jas → .. → gau + jas ecoya gAvas "cows"

This rule does not work before am (exception automzasoH).

This rule will not work on the go- that was replaced with gu by gostriyoru.

Ther rule does work on the nounbase dyo- "sky", even though it shouldn't.


Nal uttamo vA ONPANINI 71091
First person Nal (is Nit) optionally.mmmmmmmmm glosses glosses ^ C- 728 tiG

By parasmaipadAnANNal;a, both tip and mip turn into Nal. This label N lengthens nexttolast a --

pac + liT tippac + Nal ata::upa pAc + Nalpapac + apapAca "he cooked"

pac + liT mippac + Nal ata::upa pAc + Nalpapac + apapAca "I cooked"

This exception says we may erase the N label of first-person Nal. If we choose to do that, ata::upa won't work. So we may also say --

pac + liT mippac + Nal → pac + al → .. → papaca "I cooked"

With the "he cooked", however, the ata::upa is compulsory.


sakhyur asambuddhau ONPANINI 71092
Non- calling ( strong is Nit) after sakhi-.mmmmmmmmm glosses glosses ^ M- C+ 729 sup

Because of rule acoJNiti, saying that "the affix is Nit" is just a short way of saying "we replace sakhi with sakhai".

All examples --

sakhi- + au → sakhi- + Nau acoJNiti sakhai + Nau ecoya sakhAyau "two friends"

sakhi- + jassakhai- + jassakhAyaH "friends"

sakhi- + amsakhai- + amsakhAyam "friend"

sakhi- m + @calling hrasvasyaguNaH sakhes eGhra sakhe "hey friend"


anaG sau ONPANINI 71093
( sakhi-) to anaG before (non- calling) su.mmmmmmmmm glosses glosses ^ M+ C+ 730

This "to an(aG) before su" rule is just short for " sakhi- m plus normal su adds up to sakhA". So a male would say --

sakhA ca te bhaviSyAmi "and I will be your friend"

The feminine of sakhi- gets GI and works normally --

sakhI ca te bhaviSyAmi "and I will be your sister-in-law"

sakhi- m + su → sakhanssakhA "friend" ( by the same rules that make rAjA )

sakhi- f + susakhi- + GI + s halGyA sakhI "female friend"

sakhi m + @callingsakhessakhe "hey friend, hey brother-in-law"

sakhi f + callingsakhI + s → .. → sakhi "hey friend, hey sister-in-law"


Rd;uzanas; purudaMso;'nehasAM ca ONPANINI 71094
Replace R-enders, uzanas-, purudaMsas- and anehas- (with an before non- calling su).mmmmmmmmm glosses glosses ^ M- C+ 731 sup

We can split this rule into two pieces --

(1) The final R of a nonneuter nounbase merges with su into A (exception to RtoGi, that would have made ar) --

pitR m + su → pit + an + spitA "father"

mAtR f + su → mAtAnsmAtA "mother" ( same rules as pitA )

(2) The as of uzanas-, purudaMsas- and anehas- merges with su into A (exception to atvasa, that would have made As) --

uzanas + su → uzana + an + suzanA "the poet uzanas" ( same rules as pitA )

uzanA was a sage and poet of old. He is mentioned in the BHG -- kavInAm uzanA kaviH "of the poets I'm uzanas". Notice the lack of H.

vistArayitR prAjJAyAs dhyAnam "meditation is an expander of wisdom"


catur;anaDuhor Am udAttaH ONPANINI 71098
catur- and anaDuh- get Am' (before strong).mmmmmmmmm glosses glosses ^ C+ 732

This A(m) is a mitaugment. So we add A after the u --

catur- m + jas → catu + A + r + jas ikoyaNaci catvAras "four"

catur n + zi → catuAri ikoyaNaci catvAri "four"

anaDuh- m + jas → anaDuh + Am' + jas midaco anaDuAhas ikoyaNaci anadvAhas "oxen, bulls"

anaDuh- m + su → anaDuAh + s ikoyaNaci anaDvAhs sAvanaDuhaH anaDvAnhs halGyA anadvAnh saMyogAnta anaDvAn "ox, bull"


RRta:: id dhAtoH ONPANINI 71100
RR of root to ir.mmmmmmmmm glosses glosses ^ M- C+ 733

Roots with RR never keep their RR, it always changes into something else. If no other rule changes a rootfinal RR, this rule will change it into ir.

As in --

gRR + za + tip → girati

kRR + za + tip → kirati

vi + stRR + kta radAbhyAnni vi + stRR + na → vistirna halica vistIrna- Natvam vistIrNa- "extended"

See also upadhAyAzca below.


upadhAyAz ca ONPANINI 71101
nexttolast RR to ir.mmmmmmmmm glosses glosses ^ C+ 734

The roots with RR never keep their RR -- they are actually ir or ur roots, but they are listed with RR so that some special rules work on them.

This rule will change a nextolast RR into ir, unless some other rule changes the RR into something else. Example --

kRRt + Nic + zap + tip → kirt + Nic + zap + tip halica kIrt + Nic + zap + tipkIrtayati "praises"


ud oSThya-pUrvasya ONPANINI 71102
(RR to) ur after root labial.mmmmmmmmm glosses glosses ^ C+ 735

Exception to qta::iddhAtoH and upadhAyAzca, that would have made ir.

The labials are six: the five pu, and v. Examples with pu --

pRR + kta → pur + kta radA pur + kna halica pUrna Natvam pUrNa- "filled, full"

pRR + san → .. → pupUrSati "wants to fill"

Example with v --

vRR + sanvuvUrSati


sici vRddhiH parasmaipadeSu ONPANINI 72001
ik to vRddhi before sic that is before flat.mmmmmmmmm glosses glosses ^ M- C+ 736 luG

Most roots get compulsory vRddhi replacement before sic before flat -- either by this rule if they have ik, or by the next two rules if they are halanta. But this won't happen before iSic, see exception neTi below.

Example. Here R is an ik, and is before a sic that is before tip, a flat.

kR + luG tip luGlaG akR + tipakR + t'''akR + sic + t → akAr + s + takAr + s + It''' kric akArSIt "he made"


ato r;l'-.Antasya ONPANINI 72002
a of r-l-ender root to vRddhi (before sic that is before flat).mmmmmmmmm glosses glosses ^ 737 luG

Examples --

car + luG tip → .. → a + car + iSic + t''' astisicopRkte acar + iS + It iTa::ITi acarIt → acArIt



This debars the option of atohalA.


vada;vraja;hal-antasy/AcaH ONPANINI 72003
vad vraj and hal-enders to ( vRddhi before sic that is before flat).mmmmmmmmm glosses glosses ^ C+ 738 luG

Examples with aniT roots --

bhid + luG tip → .. → abhid + sic + t''' → abhaid + s + tabhaid + s + It''' kharica abhaitsIt "he broke"

chid + luG tip → .. → acchaitsIt "he splitted"

pac + luG tip → .. → apAkSIt "he cooked"

Examples with seT roots --

vad + luG tip → .. → avad + sic + t''' → avAd + s + It'''avAd + iSic + It iTa::ITi avAdIt "said"

vraj + luG tip → .. → avrAjIt ( same steps )

raJj + luG tip → .. → arAGkSIt "he colored"


n/eTi ONPANINI 72004
But not before iSic.mmmmmmmmm glosses glosses ^ C+ 739 luG iT

The previous rule won't work if the sic got iT, that is, won't work after seT roots.

Therefore, before iSic, a hal-ender will not have its vowel replaced with vRddhi. But puganta will work if it can. Examples --

div + luG jhi → .. → adiv + sic + jusadiv + iSic + us puganta adeviSus "they gambled"

div + luG tip → .. → adiv + iSic + tadiv + iSic + It''' puganta adev + iS + It iTa::ITi adevIt "he gambled"

kuS + luG tip → .. → akoSIt "he pulled out"

Back to iS-aorist.

bhid + luG tip → .. → abhaitsIt "he broke"

lU + luG tip luGlaG alU + tipalU + t'''alU + sic + t sicivRddhiH alau + s + talau + s + It''' iTa::ITi alau + It ecoya alAvIt "he reaped"


ato halAder laghoH ONPANINI 72007
Light a of a consonant-starter (to A optionally before iSic before flat).mmmmmmmmm glosses glosses ^ C+ 740 luG

Example. kaN and raN are consonant-starters, and get iSic, and their a are light (that is, followed by one consonant). So this rule may work --

kaN + luG tip → .. → akaN + iSic + t''' → akAN + iSic + takAn + iSic + It''' iTa::ITi akANIt "he shrunk"

raN + luG tip → .. → arANit "he was pleased"

Alternatively akaNIt, araNIt.

71027 yuSmad and asmad ←

chunk 38: 71078 /zatR with /num

→ 72009 /iT rules