do.

From Proto-Italic *didō, from Proto-Indo-European *dédeh₃ti, from the root *deh₃- (“give”). The reduplication was lost in Latin in the present tense, but is preserved in the perfect tense and in other Italic languages. A root aorist (from Proto-Indo-European *déh₃t) is preserved in Venetic 𐌆𐌏𐌕𐌏 (doto); the other Italic perfect forms reflect a reduplicated stative, *dedai. However, the root aorist possibly served as the source of the Latin present forms. The form duim may reflect earlier Proto-Italic *dū-ī-, from the optative of an older Proto-Indo-European aorist *déwh₃-t. The stem of this aorist perhaps derives the metathesis of *dh₃u-, itself perhaps derives from the root *deh₃- with a u-extension. Sihler, however, dismisses the proposed root extension as "ad hoc" and arguing that the proposed cognates, δοϝέναι (dowénai) and दावने (dāváne), have more plausible explanations within their own branches, thus leaving the Italic forms isolated. Another theory, advanced by the linguist Jay H. Jasanoff, suggests that the form derives from Proto-Italic *dōwīm, itself from Proto-Indo-European *déh₃-ih₁-m̥, with a glide inserted to break the hiatus. The term escaped the effects of Thurneysen-Havet's Law due to the root accent. Cognates include Ancient Greek δίδωμι (dídōmi), Sanskrit ददाति (dádāti), Old Persian 𐎭𐎭𐎠𐎬𐎢𐎺 (d-d-a-tu-u-v). The derivatives of dō are not always easy to distinguish from those of -dō (“put”) < *dʰeh₁-.