For Electrum-style mnemonic seeds, words start with a unique prefix, usually the first 3 or 4 letters, so that users can change the rest of the word. But what would happen with languages that are written from right to left, like Arabic, Persian and Hebrew? Would the code expect the end of the word to be unique? And would it be difficult to change the system to accommodate right-to-left scripts?
At the moment, the GUI has been translated into Arabic and Hebrew, but we don't have any word lists for mnemonic seeds in right-to-left languages.