Dude, that is a topic I spent an entire semester on in psycholinguistics, and wrote a paper on(focusing primarily on how bilinguals choose the proper language) and I still can't fully answer that. It's an insanely complicated subject that covers multiple aspects ranging from word and language choice to the encoding of those words into phonemic chunks and subsequently transferring those into commands for movement of articulators. For just about every step, there's considerable debate and competing models for how things work and are stored or encoded. That's not even getting into issues of executive control and function, and the homunculus problem(the idea that most models which posit free will or significant executive control outside of automaticity have difficulty explaining WHAT EXECUTIVE CONTROL IS without making it sound like a little man inside your head pulling strings). No one is going to read a book to figure out your argument. Especially not a book which uses nothing but cheesy and ultimately incorrect metaphors for the mechanism of speech