It occurred to me the other day that if ハ行転呼 had affected all applicable environments without exception, 母 /haha/ (or I guess properly it was /ɸaɸa/, right?)should have become /hawa/.
The Japanese wiki page for ハ行転呼 makes note of this and kind of handwaves it away by saying it's thought the relationship between 父/爺 and 母/婆 influenced 母 enough to keep its pronunciation from undergoing ハ行転呼, but there's no citation and it kinda just sounds to me like a "we have no clue" answer.
特殊な例として「はは」(母)のように一旦「はわ」に変化したのち、再び「はは」に戻ったものもある。これは「ちち」(父)、「ぢぢ」(爺)に対応する「はは」(母)、「ばば」(婆)の体系を維持しようとして引き起こされた変化と考えられる。
It does say 「一旦『はわ』に変化したのち」, so /haha/ (/ɸaɸa/?) did become /hawa/ before changing (back?) to /haha/. Is this maybe a case of spelling pronunciation or a conscious effort on (educated?) speakers' part to pronounce 母 in a deliberate, different way? I'm thinking like how some English speakers pronounce the "t" in often both because of its spelling and relationship to oft.
Since I couldn't find citations or sources, I guess my question is basically: Did 母 undergo ハ行転呼 and then change (back) to /haha/? Or was it an exception to the rule from the get-go?
(And now that I'm thinking about it, I wonder if there are dialects that pronounce it /hawa/, or if there were until relatively recently.)