The vowel exchange "a" to "o"

[Battler]

Sound changes in the middle of the word are fairly common. In Slovenian, there's the e deletion: mesec -> mesca, koper -> kopra, etc. It also has worth doublets that exists with both e and a - mehek and mehak, močen and močan, etc. Also words which have two different plural genitics, eg. ovca -> ovc and ovac. Also, verbs like odeti (to cover) vs. oditi (to leave). Also biti which, with different conjugations, can mean both to be and to beat. In Freising Monuments era this was likely byti vs. biti. And pre-standardization, there was probably even more such variation.

Japanese has inter-word consonant voicing (rendaku) - koi + hito = koibito, ko + shishi = kojishi, etc. It also has godan (yodan in medieval Japanese) verbs, where the vowel at the end changes, eg. yomu (to read) -> yomanai (not to read, yomanu in medieval Japanese). Medieval Japanese even had verb form pairs that only differe in a vowel, eg. yomanu (not to read) vs. yominu (to have read) (some expert in classical Japanese please correct me if I'm wrong). It even has left-overs of ancient vowel harmony: kaze + hana = kazahana (strangely not kazabana - perhaps z being voiced inhibited the rendaku?).

Dialectal forms also come to mind, eg. an Italian-speaking writer may alternate between mb and nb depending on if a standard or Venetian form came first to his mind. The same goes for mp vs. np, eg. forever would be sempre in standard Italian (and in Tuscan) but senpre in Venetian. Also ar vs. er confusion - eg. libertà in standard/Tuscan, but libartà or even łibartà in Venetian. Then there would be the thing of eg. writing facto in Latin set phrases but fatto in Italian sentences, etc.

In German, there's umlaut, eg. Baum -> Bäume. English has something similar with foot -> feet. Italian has something similar with verb - vedi (though seest) vs. vidi (I saw).