deriv LSK ETT STT aSTA ALPH OLDHOMEPAGE NEWHOMEPAGE
Nearly all /anekAc feminine bases ending in I ई end in /GI and are therefore /nadI. Even when no one knows what masculine nounbase was the /GI upposed to be added to. Example. The base **vRkI- वृकीॱ is clearly /nadI because it obviously comes from **vRka- वृकॱ —
**vRka- वृकॱ m + /GI → **vRkI- वृकीॱ f by jAterastrIv...
But the base
nadI- नदीॱ f "river"
Does not appear to come from adding /GI to anything. Yet, grammarians deem it to have been formed by adding /GI after some "nadaT" base of the same meaning.
The /GI-less feminine I ई-enders are:
(1) Intrinsically feminine /ekAc bases ending in I ई, such as **zrI- श्रीॱ, and their compounds.
(2) Bases like /grAmaNI-, that end in I ई rootnouns like nI- नीॱ "leader".
(3) Optionally, the base /lakSmI-.
Words that have /GI lose /su by halGyAbbhyodIrgh... because they have /GI, not because they are /nadI —
**nadI- नदीॱ f + /su → **nadI नदी
/lakSmI- f + /su → **lakSmI लक्ष्मी if we think it has /GI
Those that have no /GI keep their /su —
**zrI- श्रीॱ + /su → !**zrIs श्रीस् "shine, glory, holiness, goddess of wealth"
/strI- however loses /su (by !! ) —
**strI- स्त्रीॱ + /su → !**strI स्त्री " s woman"
Therefore U ऊ-enders, even if /nadI, keep the /su —
**vadhU- वधूॱ f + /su → !**vadhUs वधूस्
So do the non-/nadI ending in I ई, namely the zrI श्री-class, the /grAmaNI--class, and optionally **lakSmI- लक्ष्मीॱ —
**zrI- श्रीॱ + /su → **zrIs श्रीस्
/lakSmI- f + /su → **lakSmIs लक्ष्मीस् if we think it has no /GI
**grAmaNI- ग्रामणीॱ + /su → **grAmaNIs ग्रामणीस्
**khalapU- खलपूॱ mf + /su → **khalapUs खलपूस्