𑀓𑀟𑁆𑀠𑀇

Bloch and Hoernlé already note that the cluster -ḍḍh- has seemingly undergone irregular voicing; the expected form is *kaṭṭhaï. Other examples of such unexpected voicing are Prakrit 𑀤𑀸𑀠𑀸 (dāḍhā, “beard”), Ashokan Prakrit 𑀘𑀖𑀢𑀺 (caghati, “to be able to”); among Pali scholars this phenomenon has usually been thought of as a dialectal feature restricted to the East (while in this case it is universal), leading to Lüders and Turner rejecting the etymology.^[5][6] Tedesco pushes back and argues for generalisation of a dialectal development citing parallels like Prakrit 𑀧𑀠𑀫 (paḍhama, “first”) (but note that this is actually expected voicing), and is supported by von Hinüber and Oberlies.^[7][8]

Some scholars attempt to sidestep Sanskrit and trace the word directly to Proto-Indo-European. Gray claims derivation from an extension of Proto-Indo-European *kelH- (“to rise, be tall; hill”), with cognates such as English hale (“to pull, drag”), but this etymology seems unjustified given the lack of any cognates with the same extension.^[9] Bloomfield suggests -d- extension of some Indo-European root which gives a retroflex in Sanskrit when preceded by a sibilant (as in Proto-Indo-European *h₂eys-d- (“to laud”) > Sanskrit ईड् (īḍ)), but no semantically plausible root is to be found.^[10]

Compare also Kalami [script needed] (gaḍh, “pull!”), Phalura gaḍíi (“to pull”), Mahasu Pahari गढ़नों (gaṛhnõ), गड़नों (gaṛnõ, “to take out”).^[11] Tedesco supposes these are later developments and not indicative of a substrate source.