When kartarizap won't work.mmmmmmmmm ^ M- C+ 1306

The roots in the lukclass get nothing, and the roots in classes three to nine get other affixes, such as zlu zyan znu.

The roots in the zapclass get zap.

The sanAdyanta roots also get zap.

When a verb has zap inside, typing it into inria reader will show a "[1]", meaning that it contains a zapclass root followed by zap.

ithalmmmmmmmmm ^ M- C+ 1307

ithal is the thal that got iT in front.

Ordinarily, thal gets iT because of kRsRbhRvR. However, there are plenty of exception rules (such as acastAsvat) that make thal not get iT before certan roots, or get it optionally before others.

My advice: don't waste effort on learning those rules, because thal has been largely out of use since before the epics. Knowing that Attha means "you say" and vettha "you know" should be enough for most freshmen.

aorist typesmmmmmmmmm ^ C+ 1308 luG

When you type akArsIt into inria reader, it says --

" kR aor [4] ac sg 3 ".

kR is the root, aor means that it got luG, and the [4] means that the sic affix was added to the root, and "ac sg 3" is short for " tip". So, the word was formed this way --

kR + luG → kR tip → kR + sic + tip akArSIt "he made"

Roots before luG always get some affix. The [4] tells us which affix it got. So this [4] has nothing to do which the [4] that is not after "aor", such as nRtyati "nRt pr [4]", which means that the root got zyan.

Some examples --

aor [1] ac -- adAt "he gave" -- here rule gAti;sthA worked and the root got no affix.

aor [1] ps -- akAri "it was made" -- root got ciN

aor [2] -- adyutat "it sparkled" -- aG

aor [3] -- apIpacat "he made them cook" -- caG

aor [4] -- akArSIt "he made" -- sic

aor [5] -- acAriSam "I moved" -- iSic

aor [6] -- ayAsiSus "they went" -- siS

aor [7] -- advikSat "he hated" -- ksa

Some roots can get more than one of these.

zatR getting num before zI GI.mmmmmmmmm ^ M- C+ 1309

zatR being an ugit, it should ordinarily get num before strong only, by ugidacA. However, because of exception rules, it sometimes gets num also before the weak affixes GI and zI.

More in detail, the zatR that is before zI GI gets num --

compulsorily after zap zyan (pacantI, see zapzyanornityam).

optionally after za and after the lukclass that end in A (kSipatI kSipantI, see AcchI)

never elsewhere (dviSatI, dadatI, zRNvatI, bhuJjatI, kurvatI, krINatI etc.)

Types of compounds.mmmmmmmmm ^ M- C+ 1310

dvandva

tatpuruSa

longhorn (bahuvrIhi)

unchanging compound (avyayIbhAva)

Sanskrit compound - Wikipedia

mAtrA theory.mmmmmmmmm ^ C+ 1311

A mAtrA "measure" is a unit of time, roughly one sixth of a second. Or a bit less, if you talk fast, or a bit more, if you talk slow. Just talk at your own speed, but keep always the same speed.

The mAtrA theory says that ---

A consonant should last half a mAtrA

a double consonant should last a mAtrA

a short should last a mAtrA

a long should last two mAtre,

and an extralong should last three mAtrAH.

veda chanters try to follow those rules as closely as possible when reciting. Anything short of that is unprofessional.

These are not pANini rules, but recitation guidelines. So, when you call someone from afar applying rule dUrAddhUteca, feel free to make your extralong as long as you feel like, even until you run out of breath.

About " beforepause ".mmmmmmmmm ^ C+ 1312

Sometimes I will write things like --

azva- + su → azvas → azvaH ( beforepause )

idam- + bhis a + bhis →bahuvacanejhalyet ebhis → ebhiH ( beforepause )

azva + Gasi → azva + At → azvAt →jhalAJjazonte azvAd → azvAt ( beforepause )

This means that the rules make azvas ebhis azvAd ALWAYS, no matter what comes next, but when a pause comes next, we must replace s d with H t because some rules say so.

jhalAJjazonte changes most consonants into one of b g D d

sasaju changes s into r

kharava changes r into H

vAvasAne changes b g D d into p k T t

Everything else ( eg u n m G) stays unchanged before pause.

apANinIyanM na prayujyetammmmmmmmm ^ M- C+ 1313

This old saying translates into --

"never go against pANini"

or, using Frank Herbert's wording --

"the forms must be obeyed".

In spite of this saying, the agreement of later grammarians overrides what pANini says. As for instance, according to pANini the accent must be used in all circumstances, in the laukika too, not only when reciting the vedas. But in spite of that, Sanskrit has been spoken without the accent described by pAnini since forever.

I use these jargon words sometimes --

When I say that something is PC (Paninically Correct), that it rox, or that it is kosher, I mean that it agrees with what pANini says.

When I say that is sux, I mean it doesn't.

to save spacemmmmmmmmm ^ M- C+ 1314

When I say to save space , I mean to save time. As students were supposed to memorize all the rules, pANini made the time needed to chant all of them as short as possible. In this video, you can hear all of them sung in less than four hours --

The Complete Sutrapath

An old proverb goes --

ardhamAtrAlAghavena putrotsavam manyante vaiyAkaraNAH

"shaving off half a mAtrA from a sUtra makes a grammarian happier than his son's Bar Mitzvah"

This means that grammarians go to extreme lengths to make their rules as short as possible. Example. Rule prathamayoHpUrvasavarNaH might as well have been worded aujaszasAmpUrvasavarNaH, which would have been clearer and shorter to explain. Yet pANini chose the shorter form prathamayoHpUrvasavarNaH.

Rules trickle words down.mmmmmmmmm ^ M+ C+ 1315

The words in a rule will trickle down to the next rule, unless they have a good reason not to.

I shamelessly copied the following explanation from learnsanskrit dot org (imitation is the sincerest form of flattery). Suppose three rules look like this --

43 give a cow to Bill

44 a blanket to John

45 a shovel

Then they must be understood to mean --

43 give a cow to Bill

44 (give) a blanket to John

45 (give) a shovel (to John)

Here, the word "give" trickled down from 43 to 44 and 45, and the word "to John" from 44 trickled down to 45.

John, however, gets no cow, because the word cow and the word blanket both have second ending (in the supposed Sanskrit original of these), so the presence of the word blanket prevents the word cow from trickling. (If John were to get cow and blanket, rule 44 would have been "a blanket too to John".) Also, Bill gets no blanket, because the "to John" of line 44 prevents the "to Bill" of 43 from trickling down.

Now a Sanskrit example. Rules 34102 34103 34105 34106 literally say --

liG gets sIyuT

When flat, gets yAsuT and is Git

jha to ran

iT' to a'

Yet, because words trickle down, these rules actually mean --

liG gets sIyuT

But when ( liG) is flat, gets yAsuT and is Git

( liG) jha to ran

( liG) iT' to a'

Here the word liGas of the first rule has trickled to the next three rules. Yet the word sIyuT from the first did not fall down to the second, because the second has yAsuT (both sIyuT and yAsuT carry first ending).