Need? No. This isn't about the receivers in this situation. Other people aren't waiting for your permissions, and they don't need to be told that your opinio

Need? No. This isn't about the receivers in this situation. Other people aren't waiting for your permissions, and they don't need to be told that your opinion is your opinion. This is about the way the sender communicates their own internal understanding of their belief/knowledge states. If other people never hear someone linguistically marking their belief states with reliability indicators (even when their statements clearly need them by common phrasing norms), then the listeners will have much less reason to believe the speaker uses any reliability metrics internally, and that is just sloppy thinking. This is, I believe, the reasoning behind listeners believing that speakers who never use something like "I think" are "know it all's" who take their own opinions as gospel. If you want people to know your internal knowledge representations carry proper markers of fallibility, communicate that part of the representations in your words. That is the common norm, and other people aren't mind readers to know you mean different. Also, listeners can make good use of reliability markers from speakers, and so would like them. Such markers can communicate helpful information (close or distant source, checked validity or not, fits well with other knowns or not, reasoning from one person or checked over by many), so why leave them out? Inclusion costs so little, and omission risks unnecessary confusion. This is basic communication practice; if you don't mean to communicate to others that you consider yourself very certain on some matter, then you shouldn't send the type of messages to others that you know they will interpret as you saying you are very certain on some matter.