deriv SD cv (1146) ashtadhyayi.com hei.de L 1146 ETT STT a 8.2.9 ALPH OLDHOMEPAGE NEWHOMEPAGE

m;Ad upadhAyAz ca mator vo 'yavAdibhyaH

मादुपधायाश्च मतोर्वो ऽयवादिभ्यः ONPANINI 82009

After m म् a अ A आ that are nexttolast or last, replace /matup with v व्, unless after the yavAdi यवादि.

After last A आ a अ —

vRkSavat- वृक्षवत् "having trees"
mAlAvat- मालावत् "having garlands"

After last m म् —

zamvat शम्वत् "prosperous"
kimvat किम्वत् "having what?"

After nexttolast m म् —

dADimIvat दाडिमीवत् "having pomegranate trees"

After nexttolast a अ —

yazas- यशस् n + /matup → !**yazasvat- यशस्वत् mn "abundant in fame", "famous"

saras- सरस् n + /matup → **sarasvat- सरस्वत् mn "full of ponds or lakes"

As these are /ugit, the f gets /GI —

**sarasvat- सरस्वत् + /GI → **sarasvatI- सरस्वतीॱ

Why "not after yava यव-class"?

garutI yasya गरुती यस्य → !**garutmat- गरुत्मत् "that has wings"

!**yavamat- यवमत् "mixed with barley"

!**bhumimat- भुमिमत् "landowner"

The yavAdi यवादि (bhUmi- भूमिॱ garut- गरुत् ...) are the exceptions to this rule and the next. Don't bother memorizing the yavAdi यवादि list, it is an /AkRtigaNa. So, whenever you hear words like yavamAn यवमान् bhumimAn भुमिमान् "landowner", that flout this rule, or garutmAn गरुत्मान्, that flouts jhayaH >, they are yavAdi यवादि.

Doesn't svAdiSvasarv... make into a /pada every base that is before any /valAdi /taddhita, including /matup and its synonims? If so, why don't we get yazovat- and sarovat, as in yazodA- यशोदाॱ ?

We don't get that because of the exception tasaumatvarth.... After all, /matup has the sense of /matup.