- The former kere part has multiple possible derivations:
- Auxiliary verb maji conjugates almost identically to a classical -ku adjective, so 1 and 2 above can be broadly viewed together. This theory is somewhat flawed, in that the regular realis form of classical -ku adjectives would be not -kere, but rather -kare (as the fusion of adverbial ending -ku + the realis form are of verb ある aru “to be”). There are historical examples of -ku + are fusing into -kere, but these appear to be restricted to early modern Japanese in the Kyoto area, while keredomo appears in the late middle Japanese period and was not restricted to Kyoto.
- The realis form of past recollective auxiliary verb keri was likewise formed by fusion, combining ki with the realis form are of existence verb ある (aru, “to be”). There are other documented examples of the diphthong ia changing into e, so the shift from ki + are to kere would not be unexpected. Some theories point towards the ki element being the past recollective auxiliary verb ki, but this had no known continuative form (for joining to following verbs). Other theories suggest that this was the continuative stem form ki of Old Japanese verb 来 (ku, “to come”), which would fit grammatically, with the compound presumably meaning “having come to be, having come into being”. Compare English be that as it may.
- The latter domo part was itself originally a compound of ど (do, conjunctive particle introducing a contrary condition) + も (mo, “even, also”, particle). The do conjunctive particle grammatically requires that the preceding conjugation be in the realis form, and this is indeed the form used for the preceding kere portion of keredomo.
けれども (rōmaji keredomo)
- それでも (sore de mo)