deriv SD cv (211) ashtadhyayi.com hei.de L 211 ETT STT a 1.1.63 ALPH OLDHOMEPAGE NEWHOMEPAGE

na lumatA 'Ggasya

न लुमता ऽङ्गस्य ONPANINI 11063

Rule < pratyayalop... won't affect the stem of /luk /zlu /lup.

We can look at this as a lookahead rule —

Affixes that will be replaced with /luk /zlu /lup, do not cause the stem-changes they would ordinarily cause.

And (though it is not always a good idea) we may also think of this rule as being a rollback rule —

When an affix is replaced with /luk /zlu /lup, the stem-changes that it already caused are rolled back.

Let's see an example.

The adding of /zap after a root ordinarily triggers gbH. However, adding a /zap that will afterwards be replaced by /luk will not have that effect —

dviS द्विष् + /tas
→ dviS द्विष् + /zap + /tas by kartarizap
→ (now pugantalagh... does not work because this rule debars any stem changes before a /zap that will become /luk afterwards)
→ dviS द्विष् + /luk + /tas by adiprabhRt...
→ dviSTas द्विष्टस् by STunASTuH
→ **dviSTas द्विष्टस् "they2 hate"

We can also write the same operations as —

dviS द्विष् + /tas
→ dviS द्विष् + /zap + /tas by kartarizap
→ dveS द्वेष् + /zap + /tas by gbH
→ dveS द्वेष् + /luk + /tas by adiprabhRt...
→ dviS द्विष् + /luk + /tas by this rule, which tells us to roll back the e ए change from gbH above
→ dviSTas द्विष्टस् by STunASTuH
→ **dviSTas द्विष्टस् "they2 hate"

However

Why do we say "of /luk /zlu /lup"?

The stem of a /lopa is not affected by this exception, so the stem-changes caused by an affix stay when that affix is replaced with /lopa —

**rAjan- राजन् + /su
→ rAjAn- राजान् + /su, lengthening a अ by sarvanAmasth...
→ rAjAn- राजान् + /lopa, losing s स् by halGyAbbhyodIrgh...
→ (here the lengthening above is not rolled back because there is no this rule)
→ **rAjA राजा by other rules