Notes :Several different stems that may or may not go back to a single PDR root. The voiceless variant in PKG and Konda shows that the root could be monosyllabic as well (unless this is a regular development from a stem like *CVC-ú-). // Present in 4/6 branches.
Notes :SDR languages show significant evidence for two bases: *aḍú- ("be near to, approach") and *aḍái- ("reach, arrive, obtain"). Kannada and Tamil also have a monosyllabic stem *aṭ- "to put, place", but this is possibly the result of contamination with *anṭ- "to join, stick, paste". Cf. also *aḍ-[úŋ]- "to be piled up, arranged" which may be a historical derivative of the same root. PNDR *aṭ-g- < *aḍV́-k-. // Present in 4/6 branches.
Notes :The root is only found in Central Dravidian languages, but cannot be successfully derived from any prototype or traced back as a borrowing. // Present in 2/6 branches.
Notes :PKG *av-paḍ- < *aga-paḍ- (cf. Tamil aka-ppaṭu etc.). The root does not appear to have been preserved in its pure form outside South Dravidian. Inclusion of Telugu akku "breast" looks dubious since that form points to a monosyllabic *ak- rather than bisyllabic *agá-. // Present in 3/6 branches.
Notes :South Dravidian and Telugu languages mostly agree in displaying a bisyllabic version of the stem. It is unclear how this ties in with Malto agare, where -ar- is a typical Malto suffix. Perhaps the Malto form < *agl-are? // Present in 3/6 branches.
Notes :? Cf. *agasti "pea-tree" (the semantic connection is unclear, but phonetically the attested forms are practically identical). // Present in 2/6 branches.
Notes :A rather strange trisyllabic root with a peculiar inlaut cluster (possibly of some obscure non-Dravidian origin, although Nostratic parallels are also available). // Present in 2/6 branches.
Notes :A clearly bisyllabic verbal stem. The NDR forms present some trouble, but it is possible that *-rq- is here a regular development < *-ṛg- < *-gǝṛ-. // Present in 4/6 branches.
Notes :The geminated *-gg- ("original voiced") consonant is reconstructed on the basis of Tulu agge. All the other SDR forms reflect a variant *agái, possibly reflecting the original verbal form (i. e. PDR *ággai "shoot, sprout" vs. *agái "to flourish, sprout"). It is unclear how the Kurukh form ties into it and whether the -u- in it is relevant; if so, it is not excluded that the original form contained a cluster (*ágbai?). // Present in 2/6 branches.
Notes :The word was used with pronominal prefixes in PDR: *tāy 'his, her mother', etc. (some of these forms are innovative). The vowel was presumably long only in prefixed forms, and later was analogized in such languages as Tamil (āy 'mother').
Proto-Dravidian :*ája-
Meaning :an elder relative (father's brother; grandfather; father)
Notes :A widespread Dravidian root, although the basic meaning 'father' should probably be deemed secondary due to PDR already possessing two more likely candidates: *ápa- (coll.) and *-and_- (hon.). Gondwan vocalism is obscured by possible traces of former pronominal prefixes. // Present in 4/6 branches.
Notes :The root forms several PDR stems; the most common is *ákā~*ákai "elder sister", but a PGn stem in *-o is also available, denoting a male rather than female relative. // Present in 4/6 branches.