*tēkaną

From an apparent Proto-Indo-European *deh₁g- (‘to touch’), with the only formal cognate being Tocharian B täk- (‘to touch, feel (with the hand); to fetch’), ceśäṃ (“touches; takes”, 3sg.pres.). Ringe suggests that either pre-Germanic or pre-Tocharian borrowed the word from the other.

Several authors have noted the similarity of this root to Proto-Indo-European *teh₂g- (‘to touch, take (by the hand)’) (whence *þakkōną (‘to touch lightly, tap’), Latin tangō (‘I touch’), Ancient Greek τεταγών (tetagṓn), but the two cannot be reconciled by traditional means, as the "unshifted" Germanic *t is not expected to come from PIE *t in this phonetic environment, nor should Germanic *ē (as in the Gothic descendant only) have developed from PIE *eh₂, and Tocharian short *ä cannot continue PIE *eh₂ under normal conditions. However, the stop consonant in Germanic can be explained if it developed from a cluster *dg with voicing assimilation from earlier *tg. According to the hypothesis by Mottausch, the reduplicated form *te-th₂g- (as in Latin tetigī (‘have touched’), Ancient Greek τεταγών (tetagṓn) (tetagṓn)) regularly lost its laryngeal in pre-Germanic, and so the resulting *te-tg- would have required voicing assimilation to *te-dg-, which was subsequently back-formed to *de-dg- to restore the normal pattern of reduplication, resulting in the unshifted Germanic /t/. Additionally, the pre-form *deh₁g- that resulted in the long vowel in Germanic can be maintained, if it is assumed that the second stop in this pre-Germanic *dedg- underwent dissimilation to glottalic *h₁ = [ʔ]. Kroonen upholds this derivation, though he notes its great difficulties. For a partial alternative, he suggests a back-formation of *tēkaną from iterative *þakkōną (cf. his similar derivation of *slēpaną from *slapp/bōną).